Journal of Petroleum Science and Engineering 152 (2017) 675–682

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

A fully compositional model considering the effect of nanopores in tight MARK

oil reservoirs
⁎
Bicheng Yana, Yuhe Wangb, , John E. Killougha
a
Department of Petroleum Engineering, Texas A & M University, College Station, TX, USA
b
Department of Petroleum Engineering, Texas A & M University at Qatar, Doha, Qatar

A R T I C L E I N F O A BS T RAC T

Keywords: Conventional compositional simulators are usually difficult to interpret the different gas oil ratio (GOR) from
Tight oil reservoirs tight oil reservoirs, and this also indicates an unreliable prediction of ultimate hydrocarbon recovery. We realize
Compositional simulation that there are two issues related to the compositional simulation of production in tight oil reservoirs. Firstly,
Vapor-liquid equilibrium tight oil reservoirs typically exhibit extremely small matrix pore size in the order of nanometers, so the capillary
Capillary pressure
pressure between vapor and liquid phases is considerable such that the PVT of the confined fluid deviates from
Rock compaction
that of the bulk fluid with capillary pressure ignored. Secondly, during depletion process, rock compaction
causes pore space reduction and brings remarkable changes in rock properties. In this work we implement
rigorous confined fluid phase behavior calculation depending on capillary pressure and rock compaction in a
fully compositional simulator. Capillary pressure in matrix nanopores is calculated by Leverett J-function.
Further, the impact of capillarity on phase equilibrium is taken into account through modifying the stability test
and two-phase flash calculation. Dynamic rock compaction is considered in the simulator via rock compaction
tables, such that fluid mobility decreases with permeability reduction and capillary effect is simultaneously
coupled. The unique implementation in the simulator captures the dynamic behavior of rock and fluid
properties in tight oil reservoirs. Typical suppression of bubble point pressure and reduction of oil viscosity and
density is observed from our simulation results. Reservoir-scale simulation results show that this model resolves
the problem of the inconsistent GOR in tight oil production and greatly facilitates the history matching process.
The enhanced compositional simulation will ultimately improve our understanding of tight oil reservoirs and
provide better guidance for recovery prediction.

1. Introduction
The development of unconventional tight oil resources has achieved
great success through the application of horizontal well drilling and
hydraulic fracturing. As predicted by EIA (Fig. 1), tight oil production
worldwide will increase to be more than double from 2015 to 2040
(Aloulou 2016), and specifically tight oil production in United States
will increase from 4.1 million barrels/day in 2016 to 7.1 million
barrels/day in 2040. Despite of this great success, our understanding of
the production mechanisms in tight oil reservoirs is still very limited.
In tight oil resources, the matrix permeability is extremely low and
the average pore size is in the nanometer magnitude. For example, the
average pore size in Bakken shale matrix is about 10 to 50 nm in
middle Bakken region calculating based on Kozeny-Carman equation
(Nojabaei et al. 2013). Reported by Wang et al. (2012), unconventional
resources such as Bakken are usually oil-wet or intermediate-wet, and
this can be caused by the oil-affinity of organic matter in the matrix
media (Wang and Reed 2009). Fluid within such tiny pores exhibits
different properties from conventional oil reservoirs, because the
capillary pressure in those nanopores is significant and it may induce
deviation in phase behavior of tight oil reservoirs. The effect of porous
media on the phase behavior of hydrocarbon fluid was investigated in
literatures. Sigmund et al. (1973) concluded that the effect of curvature
on phase behavior is not significant except at high surface curvatures.
However, this would not happen in hydrocarbon reservoirs even with
the presence of clay particles, since those finest pores are likely to be
filled by connate water. Through implementing a capillarity equation in
phase equilibrium, Brusilovsky (1992) demonstrated that bubble point
pressure decreases but dew point pressure increases. From the aspect
of energy, Coussy (2010) stated that due to the extra energy cost
associated with surface energy effects, the difference between bulk-fluid
saturation pressure and equilibrium liquid phase pressure considering
capillary pressure increases as the pore size becomes smaller.
Further, Nojabaei et al. (2013) investigated the impact of nanopores

⁎
Corresponding author.
E-mail address: yuhe.wang@qatar.tamu.edu (Y. Wang).

http://dx.doi.org/10.1016/j.petrol.2017.01.005
Received 8 November 2016; Received in revised form 23 December 2016; Accepted 4 January 2017
Available online 05 January 2017
0920-4105/ © 2017 Published by Elsevier B.V.
B. Yan et al.
Nomenclature
Variables
fiα Fugacity of component i in phase α
FV Vapor mole fraction, fraction
K Permeability, md
Ki Component equilibrium ratio, fraction
nc Number of non-water components in the reservoir
PV Vapor phase pressure, psia
PL Liquid phase pressure, psia
Pc Capillary pressure in between liquid and vapor phases,
psia
r Pore radius in porous medium, m
R Gas universe constant, 8.314 JK−1mol −1
Rf , i Fugacity equilibrium residual of component i
R Pc Capillarity residual
Rrr Residual of Rachford-Rice Equation
T Reservoir temperature,°F
VLE Vapor liquid equilibrium
xi Molar fraction in liquid phase, fraction
Fig. 1. World tight oil production from 2015 to 2040 (Aloulou 2016).
on phase behavior of Bakken shale oil, and the results showed that
bubble-point pressure can be supressed significantly, and the same
conclusion was also reported in other literatures (Du and Chu 2012;
Honarpour et al. 2012; Pang et al. 2012). Nojabaei et al. (2014)
implemented a compositionally-extended black-oil model considering
capillarity in phase behavior, and the capillary pressure varies with
interfacial tension and pore size distribution. This approach is inno-
vative but the extended black-oil formulation is intrinsically limited in
terms of compositional characterization. Sandoval et al. (2015) devel-
oped an efficient and robust algorithm for calculating fluid phase
behavior in the presence of capillarity. They found that the bubble
point pressure suppression becomes significant when moving away
from the critical point, while the dew point pressure exhibits different
variations depending on the location of phase diagram. Rezaveisi et al.
(2015) implemented capillarity in flash and stability test in UTCOMP
simulator, but still they didn’t apply it for practical field-scale applica-
tion of tight oil reservoirs. Dong et al. (2016) coupled Peng-Robinson
Equation of State with capillary pressure equation and adsorption
theory to investigate phase behavior in cylindrical nanopores, and they
found that bubble point pressure is overestimated if neglecting
adsorption film.
Moreover, Nojabaei et al. (2013) reported that as pore pressure
decreases or effective stress increases, the tight matrix is likely to
experience a reduction of pore size. This observation can also be
inferred from history matching of production data in Bakken shale
(Nojabaei et al. 2013). In their approach, through PVT analysis they
theoretically demonstrated that the suppression of bubble point
pressure becomes even more significant when considering pore size
reduction through the depletion process, and applied this as input in a
676
Journal of Petroleum Science and Engineering 152 (2017) 675–682
yi Molar fraction in vapor phase, fraction
zi Molar fraction in the overall mixture, fraction
ϕiα Fugacity coefficient of phase α
Φ Porosity, fraction
μ Viscosity, cp
ρ Density, lb / ft 3
μiα Chemical potential of component i in phase α
σ Interfacial tension
θ Contact angle between liquid and vapor phase
Superscripts and subscripts
L Liquid phase
V Vapor phase
Abbreviations .
BHP Bottom-Hole Pressure
BIC Binary Interacting Coefficient
GOR Gas oil ratio
IMPEM Implicit pressure explicit mass
SSI Successive substitution iteration
VLE Vapor Liquid Equilibrium
commercial reservoir simulator for history matching. Therefore, a
rigorous consideration of pore size change throughout the tight oil
production processes is also very necessary for general compositional
reservoir simulation.
Motivated by those points, this work implements the impact
capillarity and rock compaction in a general fully compositional
reservoir simulator (Dean and Lo 1988) and applies it to predict tight
oil reservoir performance. Specifically capillary equation is explicitly
implemented in stability test based on Gibbs free energy approach and
vapor-liquid equilibrium (VLE). Further, matrix pore size is iteratively
calculated during pressure depletion by rock compaction tables. The
implementation is designed to rigorously consider the fluid and rock
properties variations in the nanopores. The reminder of this paper is
organized as follows. The next session illustrates the vapor-liquid
equilibrium formulation considering capillary pressure and the overall
phase behavior evaluation workflow in compositional simulation. The
third session presents our main results based on the new model.
Finally, the whole work is concluded and summarized.
2. Numerical formulation
In this work, a fully compositional reservoir simulator (Dean and Lo
1988; Tang and Zick 1993) is extended to accommodate the effect of
capillary pressure on vapor-liquid equilibrium. The simulator is
designed volume-balance formulation and it is solved by the classic
implicit pressure explicit mass (IMPEM) scheme. Since simulator
formulation itself is not the focus of this paper, more details with
regards to this can be referred to (Wong et al. 1990; Young and
Stephenson 1983). Originally the simulator is able to handle multi-
phase multi-component reservoir simulation, and the PVT modules in
the simulator include stability test based on Gibbs free energy approach
and vapor-liquid phase split calculation (Michelsen 1982a, 1982b), but
the effect of capillary pressure on phase behavior is not implemented.
In this part, we will basically introduce the newly implemented flash
and stability test algorithm considering capillarity effect in the reser-
voir simulator.
2.1. Stability test considering capillary pressure
In this part, the consideration of capillarity effect in stability test is
based on the stability criterion of Michelsen (1982a). If the original
system is stable, Eq. (1) will be valid. μi ( y) and μi (z ) are respectively the
B. Yan et al. Journal of Petroleum Science and Engineering 152 (2017) 675–682

chemical potential of the incipient phase and the original system. Table 1
nc Pore radius, permeability and capillary pressure.
F ( y) = ∑ yi [μi ( y)−μi (z )] ≥ 0 Pore Radius (nm) Premeability (md) Capillary Pressure (psi)
i =1 (1)
Based on Eq. (2), chemical potential is usually convenient to be 50 0.0070 102.20
40 0.0046 127.75
represented in terms of fugacity, given by Eqs. (3) and (4). Here it is
30 0.0027 170.33
assumed that the original phase is liquid phase and the trial phase is 20 0.0012 255.50
vapor phase. The opposite setting, where the original phase is vapor 10 0.0003 511.00
and the trial phase is liquid, is equivalent.
μi = RTln ( fi ) (2) Table 2
Rock compaction table of Bakken shale.
fi (z )=zi ϕi (z ) P L (3)
Pressure Change (psi) Permeability Reduction Ratio
fi ( y)=yi ϕi ( y) PV (4)
−5180 0.489
As a result, given system temperature T0 , Eq. (1) is transformed into −4450 0.500
Eq. (5). The last term (ln PV −ln P L ) in Eq. (5) is usually ignored because −3700 0.511
of the small capillary pressure between vapor and liquid phases. −2960 0.532
−2220 0.588
However, it is taken into account here because capillary pressure is
−1480 0.675
hypothesized to be significant in confined nanopores. −740 0.791
nc 0 1.0
F ( y)
g ( y)= = ∑ yi {ln [ fi ( yi)] − ln [ fi (zi )]}
RT0 i =1
nc
= ∑ yi {ln[ yi ϕi ( y) PV ]−ln[zi ϕi (z ) P L ]} No pressure maintenance
i =1
nc
= ∑ yi {ln yi +ln ϕi ( y)−ln zi −ln ϕi (z )+[ln PV −ln P L ]}
i =1 (5)
Eq. (6) is defined to be independent of y , and thus it can be Reservoir Compaction
calculated in advance, which is convenient in terms of implementation.
Then we have Eq. (7).
hi=ln zi +ln ϕi (z )+ln P L , (i=1, ⋯, nc ) (6)
nc
g ( y)= ∑ yi (ln yi +ln ϕi ( y)−hi +ln PV ) ≥ 0, (i=1, ⋯, nc )
i =1 (7) Greater Lower
The stationary criterion is Eq. (8). Confinement Permeability
ln yi +ln ϕi ( y)− hi +ln PV =k , (i=1, ⋯, nc )

Let Yi =e−k yi , then

(8)
+ _
ln Yi+ln ϕi ( y)− hi +ln PV =0, (i= 1, ⋯, nc ) (9)
Flow Capacity
Based on material balance Eq. (10), we can calculate yi through Eq.
(11), and this indicates that Yi can be interpreted by phase composition
moles.
Fig. 2. Impact of rock compaction and fluid confinement on flow in tight porous media.
nc nc
∑ yi =e k ∑ Yi=1 brium in vapor and liquid phases. It’s noted that those two consecutive
i =1 i =1 (10)
equations are established based on continuous bulk vapor and liquid
Yi phase separated by a curvature interface. In ultra-tight porous media
yi = nc
∑i =1 Yi (11) those two phases are not necessarily to be continuous, but it can be
shown that pressure and phase compositions in different isolated vapor
Here the solution of Eq. (9) can be used to examine the stability
nc or liquid regions are equal when the system reaches equilibrium
test. In terms of Y , this original system is stable if ∑i =1 Yi≤1 and is
nc (Bedrikovetsky 1993). Therefore, the continuous phase prerequisite is
unstable if ∑i =1 Yi >1. Eq. (9) here is solved by the combination of
assumed to be still reliable here.
Successive Substitution Iteration (SSI) and Newton-Raphson method.
R Pc=(PV −P L ) − Pc=0 (12)
2.2. Vapor-liquid two-phase split considering capillary pressure L
R f ,i = fi (T , PL, xi ) − fiV (T , PV , yi) = 0, (i=1, 2, ⋯, nc ) (13)
The consideration of capillarity effect on vapor-liquid equilibrium L L
fi =xi ϕi PL (14)
(VLE) can be readily represented by Eqs. (12) and (13). In Eq. (12), PV
and P L respectively designate vapor phase pressure and liquid phase
fiV =yi ϕiV PV (15)
pressure, and Pc represents the capillary pressure between those two
phases. Eq. (13) is the equality of fugacity in liquid and vapor phases, To close the nonlinear system in Eqs. (12) and (13), other
and the two fugacities are defined through Eqs. (14) and (15). In constraints are required. The first one is the material-balance con-
conventional flash calculation, PV is considered to be equal to P L and straint with the introduction of vapor mole fraction FV , as shown in Eq.
thus conventional VLE only includes the component fugacity equili- (16).

677
B. Yan et al. Journal of Petroleum Science and Engineering 152 (2017) 675–682

Fig. 3. Workflow for vapor-liquid equilibrium considering capillarity and rock compaction.

Table 3
Bakken oil composition data.

Component Mole Fraction Critical Pressure (psia) Critical Temperature (°F) Acentric Factor Molar Weight Parachor

C1 0.36736 667.80 -116.63 0.0130 16.04 74.8

C2 0.14885 707.80 90.09 0.0986 30.07 107.7
C3 0.09334 616.30 206.01 0.1524 44.10 151.9
C4 0.05751 550.70 305.65 0.2010 58.12 189.6
C5-C6 0.06406 461.29 415.81 0.2684 78.30 250.2
C7-C12 0.15854 363.34 593.58 0.4291 120.56 350.2
C13-C21 0.07330 249.61 872.43 0.7203 220.72 590.2
C22-C80 0.03704 190.12 1384.82 1.0159 443.518 1216.8
Reservoir temperature (°F) 240

Table 4
Bakken oil binary interaction coefficient (BIC).

C1 C2 C3 C4 C5-C6 C7-C12 C13-C21 C22-C80

C1 – 0.005 0.0035 0.0035 0.0037 0.0033 0.0033 0.0033

C2 0.0050 – 0.0031 0.0031 0.0031 0.0026 0.0026 0.0026

Other BICs are zero and not presented here for simplicity.

Table 5
Bubble point pressure of Bakken oil at 240 °F.

Pore radius (nm) Bubble point pressure (psia)

Young-Laplace Leverett J-Function

50 2766 2512
40 2761 2450
30 2751 2345
20 2732 2145
10 2641 1588
No Capillary 2788

Fig. 4. Bubble point pressure lines of Bakken oil with considering capillarity under
to the classic Rachford-Rice Equation (Rachford and Rice 1952), as
different pore radii and without considering capillarity.
shown in Eq. (19).
zi=FV yi +(1 − FV ) xi (16) nc nc nc
∑ xi= ∑ yi= ∑ zi=1
i =1 i =1 i =1 (17)
Additionally component mole fractions of each phase and the
overall mixture should sum to unity when equilibrium is reached, as
y
shown in Eq. (17). With the definition of equilibrium ratio Ki in Eq. Ki= i , (i=1, ⋯, nc )
(18), the material-balance constraints in Eq. (16) can be transformed xi (18)

678
B. Yan et al. Journal of Petroleum Science and Engineering 152 (2017) 675–682

45 0.50
Blue dash line: oil viscosity without capillarity

Oil density, lb/cu.ft

0.42

Oil viscosity, cp
Purple dash line: oil density without capillarity
40
0.34

0.26
35
0.18

30 0.10
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Pore radii, nm Pore radii, nm

(a) (b)
250
oil ρ/μ, lb/cu.ft/cp 200

150

100
Green dash line: oil ρ/μ without capillarity
50

0
0 10 20 30 40 50 60
Pore radii, nm
(c)
Fig. 5. Confined fluid properties change with pore radius at 240 °F and 1500 psia: (a) oil density; (b) oil viscosity; (c) the ratio of oil density to oil viscosity.

1.6 4500

4000

3500
1.2
Cum Oil Production, RB

3000

GOR in well, RCF/RB

2500
0.8
Oil Prod for Case 1 Oil Prod for Case 2 2000
Oil Prod for Case 3 GOR for Case 1
1500
GOR for Case 2 GOR for Case 3
0.4
1000

500

Fig. 6. 1D core model with a producer. 0.0 0

0 1 2 3 4 5
Time, days

Table 6
Fig. 7. Cumulative oil production and GOR at in-situ reservoir conditions for three
Reservoir parameters for 1D core size model.
different scenarios in the 1D model: Case 1: with no capillarity effect on VLE and no rock
compaction, Case 2 with capillarity effect on VLE but no rock compaction, Case 3 with
Model size ( ft 3) 0.5 x 164 x 0.5
capillarity effect on VLE and rock compaction.
Grid size ( ft 3 ) 0.5 x 3.28 x 0.5
Grid number 1 x 50 x 1
Porosity 0.06 Table 7
Permeability (mD ) 0.002 Reservoir parameters for the 3D tight oil reservoir model.
Initial pressure ( psia ) 6840
Producer minimum BHP ( psia ) 1500 Model size ( ft 3 ) 1000 x 1000 x 20
Producer oil rate (RB/day) 0.001 Grid number 35 x 67 x 10
Initial temperature (°F ) 240 Matrix porosity 0.06
Matrix permeability (mD ) 0.002
Fracture number 4
nc Fracture grid width ( ft) 1.0
(Ki−1) zi
Rrr = ∑ Fracture porosity 0.02
i =1
1+FV (Ki − 1) (19) Fracture conductivity (mD∙ft ) 100
Initial pressure ( psia ) 6380
Based on Eqs. (14) and (15), the definition of component equili- Producer minimum BHP ( psia ) 1900
brium ratio Ki can be further transformed into Eq. (20). Eq. (20) shows Producer oil rate (RB/day) 50
Initial temperature (°F ) 240
that the capillary pressure is embedded in this modified component
equilibrium ratio, and it can be reduced to conventional format
Ki=ΦiL /ΦVi if the capillary pressure between vapor and liquid phases is
2.3. Capillary pressure evaluation in tight porous media
ignored.

y ϕ L PL The calculations above require the evaluation of capillary pressure.

Ki= i = iV V , (i=1, ⋯, nc ) During the iteration of stability test and VLE phase split calculation,
xi ϕi P (20)
capillary pressure values are calculated using saturation results from
At a given capillary pressure, Eqs. (13) and (19) are solved through the previous iteration. The capillary pressure is fundamentally related
combining successive substitution iteration (SSI) and Newton-Raphson to the geometry of the capillary and the wettability of the porous media,
method. and it can be calculated through Young-Laplace equation, shown as Eq.

679
B. Yan et al.
Fig. 8. Tight oil reservoir with horizontal well and multiple hydraulic fractures. Confined
zone indicates the tight matrix with the consideration of capillary pressure in VLE, and
unconfined zone represents the fracture system with the consideration of bulk phase
behavior.
14000000
Oil Prod for Case 1
12000000 Oil Prod for Case 3
GOR for Case 2
Cum Oil Production, RB
10000000
8000000
6000000
4000000
2000000
0 0
0 200 400 600 800
Fig. 9. Cumulative oil production and GOR at in-situ reservoir conditions for three
different scenarios in the reservoir model: Case 1: with no capillarity effect on VLE and
no rock compaction, Case 2 with capillarity effect on VLE but no rock compaction, Case 3
with capillarity effect on VLE and rock compaction.
(21). This approach has been applied in Nojabaei et al. (2013),
Honarpour et al. (2012), and Pang et al. (2012) etc, and the pore
radius is taken into account.
2σcosθ
Pc=
r (21)
As Nojabaei et al. (2013) stated, capillary pressure calculated
through Young-Laplace equation is much lower than the experimental
measurement due to the very low interfacial tension predicted by
Macleod-Sugden correlation (Pederson and Christensen 2007). This
observation also indicates the importance of reliable capillary pressure
prediction for shale/tight rocks. Based on molecular dynamics model,
Wang et al. (2016a) successfully investigated the effect of pore size and
temperature on the surface tension and contact angle in the shale
nanopores. Yet due to the complexity of this approach, the implemen-
tation of this method in our compositional model is beyond the scope of
this work. On the other hand, Leverett J-function (Eq. (22)) is based on
measured reference capillary pressures, and the scaling factor J (S ) is
supposed to be given for a particular rock type based on lab data.
Unfortunately so far there is no solid measurement of capillary
pressure for the Bakken shale. Because the porous media investigated
here is a saturated one, the Leverett J-function approach is assumed to
be appropriate to evaluate capillary pressure here. Compared to Young-
Laplace Equation, Leverett J-function is friendly to the simulator
implementation, since pore size information is embedded in perme-
ability and porosity. The prediction of surface tension for both Young-
Laplace Equation and Leverett Leverett J-function approach are based
on Macleod-Sugden correlation (Pederson and Christensen 2007),
which is a function of phase mole fractions, phase molar densities
and component Parachor parameter.
k
Pc (S )
ϕ
J (S )=
σcosθ
The porosity of the Bakken shale is assumed to be 0.06. Since the
Kozeny-Carman equation of permeability was demonstrated to be
applicable in shale in the derivation work by Civan (2010), in our
work the corresponding permeability and pore radius can be calculated
based on Kozeny-Carman equation (Carman 1937; Civan 2010; Kozeny
1927) and Nelson’s correlation (Nelson 1994). The reference capillary
Journal of Petroleum Science and Engineering 152 (2017) 675–682
Confined pressures under different pore radii are also presented based on a
similar formation from Crain’s petrophysical handbook. Those values
are illustrated in Table 1 and are expected to be very close to practical
Unconfined situation.
2.4. Rock compaction towards nanopore space
Dynamic rock compaction is considered in the simulator through
rock compaction table. Therefore, when pore pressure changes during
2000 time-step and newton-step evolution, permeability reduction ratio is
Oil Prod for Case 2
1800 interpolated from the rock compaction table. The permeability update
GOR for Case 1
GOR for Case 3 1600 further impacts the pore radius and capillary pressure. For Bakken
shale investigated in this work, the rock compaction table is presented
1400
GOR in the well, RCF/RB
1200 in Table 2, which is based on the work of Nojabaei (2015).
1000 With the consideration of rock compaction, pore size reduces
800 during depletion, and the bubble point pressure for confined fluid will
600 be even more suppressed during production. As a result, reservoir fluid
400 is likely to stay in single phase (oil) with light components still
200 dissolved in it, and oil density and viscosity decrease and oil compres-
sibility increases such that fluid flow capacity and flow driving energy
1000 1200 1400 1600 1800 2000
Time, days increases. On the other hand, permeability decreases with rock
compaction, and thus it negatively impacts the fluid flow capacity.
Therefore, when considering rock compaction for fluid flow in tight
porous media, it induces greater confinement effect and lower perme-
ability and those two effects compete with each other, as shown in
Fig. 2.
2.5. VLE workflow considering capillarity and rock compaction
In summary, the workflow of vapor liquid equilibrium with the
consideration of capillary pressure and rock compaction is shown in
Fig. 3. At a given thermodynamic condition in a time/newton step,
through compaction table rock permeability and porosity are updated
and used to evaluate capillary pressure. Meanwhile, if previously the
fluid stays in single-phase, stability test is performed and capillarity is
considered in the calculation. If the fluid is unstable at single phase or
it is previously in two-phase, then two-phase flash calculation con-
sidering capillarity is performed. Otherwise, fluid is still single phase.
After the phase split, fluid properties and other secondary variables are
updated, and capillary pressure is calculated and is resent to stability
test and flash calculation, and new secondary variables are updated.
This process is iterated until both phase equilibrium and capillarity
equality are ensured, and then simulation gets into solving a global
non-linear system for updating primary variable(s).
3. Cases analysis
This part mainly presents the results of fluid PVT in confined
nanopores and the application into compositional simulation of tight
oil reservoirs. Fluid compositional data is based on Bakken shale oil
from Nojabaei et al. (2013), as shown in Tables 3 and 4. At the
temperature of 240 °F the saturation pressure for this fluid here is the
bubble point pressure. The fluid data in Tables 3 and 4 will be used in
all cases presented below, and the rock compaction data in Table 2 will
be used in those cases below if rock compaction is considered.
3.1. Confined fluid phase behavior
The P-T diagram of the Bakken oil is calculated through rigorous
(22)
stability test and saturation pressure calculation algorithms in session
2. The consideration of capillary pressure is evaluated through Young-
Laplace equation under different nanopore radii, and compared with
the conventional scenario without capillarity. The results are plotted in
Fig. 4. Here it shows that at a given temperature the bubble point
pressure decreases as pore size reduces. When pore radius is 10 nm,
the suppression of bubble point pressure is about 125 psia.
680
B. Yan et al.
Interestingly the suppression of the bubble point pressure becomes
weaker with the increase of temperature. This can be explained by that
the deviation between bulk-fluid and confined fluid bubble point
pressures are induced by the energy cost from surface energy effects
(Coussy 2010), and this extra cost would be counteracted by increased
molecular kinetic energy as temperature increases. Besides, the con-
finement caused by capillarity is also the interaction between fluid
molecules and pore wall. As indicated by Wu et al. (2016), when
temperature increases above a threshold value, the molecules have
sufficient kinetic energy to be less constrained by the wall attraction
field such that the wall-molecules interaction becomes minimum.
As previously discussed, capillary pressure predicted by this
approach is much lower than the actual one. Therefore, we further
evaluate capillary pressure through Leverett J-function approach for all
cases, and the results are presented in Table 5. We can see that there
are significant differences between bubble point pressures based on
Young-Laplace equation and those based on Leverett J-function. The
differences here also reveal that capillary pressure data with good
quality is very important in phase behavior prediction.
If compared to the same fluid system with unsuppressed bubble
point pressure, a suppression in bubble point pressure physically
indicates that there are more light components remaining in the liquid
oil phase under the same pressure and temperature condition. More
light components in oil phase are expected to bring a reduction of oil
density and oil viscosity. Fluid viscosity here is calculated by Lohrenz-
Bray-Clark correlation (Lohrenz et al. 1964). With capillary pressure
considered in phase behavior, Fig. 5 illustrates the oil density, oil
viscosity and their ratio changes with nanopore radius for Bakken oil at
240 °F and 1500 psia. The results are also compared with bulk fluid
properties without considering capillary pressure. In Fig. 5(a) and (b),
it clearly shows that oil density and oil viscosity decrease as nanopore
radius decreases from 50 nm to 10 nm, and the two properties are
smaller than those values when neglecting capillary pressure in phase
behavior. Further, since oil mobility is directly proportional to this
ratio, the ratio of oil density to oil viscosity is also plotted in Fig. 5(c).
We can see that as nanopore radius decreases, the ratio of oil density to
viscosity increases, and thus oil becomes much more mobile in
confined nanopore space. Besides, more light components in oil phase
also indicates that oil phase compressibility increases, which provides
more driving energy to extract liquid from tight oil reservoirs.
Molecular dynamic simulation work (Afsharpoor and Javadpour
2016; Wang et al. 2016b; Wang et al. 2016d, 2016e) also demonstrated
that fluid properties in shale nanopores are reduced and favors oil
transport in nanopores, and provided excellent support for our
reservoir-scale simulation results in this work.
3.2. 1D Core Size Model
The first example is a one dimensional core size model, as shown in
Fig. 6. Other parameters related to this model is presented in Table 6. A
producer is located at the first grid block, and it is constrained by a
minimum BHP (1500 psia) and maximum oil volume rate (0.001 RB/
day). The oil rate is very low since the model is small. The outer
boundary condition in this 1D core model is no flow boundary
condition.
Three different scenarios are simulated, respectively Case 1 with no
capillarity effect on VLE and no rock compaction, Case 2 with
capillarity effect on VLE but no rock compaction, and Case 3 with
capillarity effect on VLE and with rock compaction. The results are
plotted in Fig. 7. Fig. 7 presents the cumulative oil production and gas-
oil-ratio in the wellbore condition. Through comparison, it clearly
shows that Case 2 produces the most oil from the reservoir, while Case
1 produces the least oil from the reservoir. The comparison between
those two cases shows that capillary pressure effect on phase behavior
favors oil production in the tight oil reservoir, since oil mobility of
confined fluid increases. Case 3 considers both capillary pressure in
681
Journal of Petroleum Science and Engineering 152 (2017) 675–682
VLE and rock compaction. As expected this case produces median oil
production here. This indicates that during pressure depletion rock
compaction reduces rock permeability and it offsets the increase of
mobility caused by capillary pressure. The GOR at the in-situ reservoir
condition tells us that capillary pressure induces lower bubble point
pressure and delays gas phase appearance, and this is even more
exacerbated by further considering rock compaction. Besides, the early
zero GOR at the in-situ reservoir conditions indicates that BHP at that
period is higher than the bubble point pressure.
3.3. Fractured tight oil reservoir with multiple hydraulic fractures
Here a tight oil reservoir is perforated by a horizontal well and 4
planar hydraulic fractures traverse through the reservoir in y-direction.
The horizontal well is constrained by constant oil rate (50 RB/day) and
minimum BHP (1900 psia). The outer boundary condition in this
reservoir model is no flow boundary condition. Basic reservoir para-
meters are presented in Table 7. The schematic of the model is
illustrated in Fig. 8. Therefore, fluid flow through the reservoir follows
the sequence of matrix-fracture-well. In hydraulic fractures capillary
pressure is not considered since we assume that capillary pressure
effect is negligible and thus bulk fluid phase behavior is appropriate
here, namely unconfined zone in Fig. 8. On the other hand, in the tight
oil matrix the effect of capillary pressure on phase behavior and the
rock compaction is conditionally considered, and in Fig. 8 it is labeled
as confined zone.
Similarly, here we also compare three different scenarios in the
reservoir model, and the results at the in-situ reservoir conditions are
illustrated in Fig. 9. Here it shows that the case with capillary pressure
in VLE but no rock compaction (Case 2) has the highest cumulative oil
production, and the case with both capillarity and rock compaction
(Case 3) has the median oil production. This further demonstrates that
at reservoir scale the consideration of capillarity in phase behavior in
Case 2 facilitates the oil transport into the wellbore, while the median
oil production in Case 3 is caused by the permeability decrease due to
rock compaction. There is a decrease of in-situ GOR for cases
considering capillary pressure in phase behavior. The in-situ GOR for
the case with compaction and capillary pressure interestingly exhibits a
step-like increase and decrease back to near zero. This might be
explained by that nanopore size reduction due to rock compaction
further suppresses the bubble point pressure and thus most of the
hydrocarbon molecules flow in liquid phase in the wellbore.
4. Conclusions
In this work a fully compositional reservoir model considering the
impact of capillary pressure and rock compaction on vapor liquid
equilibrium is developed, and Leverett J-function is used to evaluate
capillary pressure for tight oil. This model is used to predict the phase
behavior in confined nanopore space and tight oil reservoir perfor-
mance. The capillary pressure effect on vapor-liquid equilibrium is
significant when pore size in the matrix is in the nanometer scale. For
oil-wet tight oil matrix, fluid confinement in the nanopores suppresses
the bubble point pressure, and results in decreases in oil density and
viscosity. This ultimately favors the oil phase transport to the wellbore
by increasing the oil mobility and more driving energy. Besides,
through our compositional model oil production in tight oil reservoirs
is demonstrated to very sensitive to rock compaction during reservoir
depletion. Rock compaction directly decreases rock permeability and
reduces fluid mobility. On the other hand, pore size reduction during
rock compaction further exacerbates bubble point pressure suppres-
sion of confined fluid in nanopores. Therefore, reservoir simulation of
tight oil reservoirs should consider both capillary pressure and rock
compaction on production.
B. Yan et al.
References
Afsharpoor, A., Javadpour, F., 2016. Liquid slip flow in a network of shale noncircular
nanopores. Fuel 180, 580–590. http://dx.doi.org/10.1016/j.fuel.2016.04.078.
Aloulou, F. 2016. World Tight Oil Production to More Than Double from 2015 to 2040.
International Energy Outlook 2016.
Bedrikovetsky, P., 1993. Mathematical Theory of Oil and Gas Recovery. Petroleum
Engineering and Development Studies. Springer Netherlands, The Netherlands,
Original edition. ISBN 978-90-481-4300-9.
Brusilovsky, A.I., 1992. Mathematical simulation of phase behavior of natural
multicomponent systems at high pressures with an equation of state. SPE Reserv.
Eng. 7 (01). http://dx.doi.org/10.2118/20180-PA.
Carman, P.C., 1937. Fluid flow through granular beds. Chem. Eng. Res. Des. 75,
S32–S48. http://dx.doi.org/10.1016/S0263-8762(97)80003-2.
Civan, F., 2010. Effective correlation of apparent gas permeability in tight porous media.
Transp. Porous Media 82 (2), 375–384. http://dx.doi.org/10.1007/s11242-009-
9432-z.
Coussy, O., 2010. Phase Transition in Porous Solids. In Mechanics and Physics of Porous
Solids. John Wiley & Sons, Ltd..
Dean, R.H., Lo, L.L., 1988. Simulations of naturally fractured reservoirs. SPE Reserv.
Eng. 3 (02). http://dx.doi.org/10.2118/14110-PA.
Dong, X., Liu, H., Hou, J., et al., 2016. Phase equilibria of confined fluids in nanopores of
tight and shale rocks considering the effect of capillary pressure and adsorption film.
Ind. Eng. Chem. Res. 55 (3), 798–811. http://dx.doi.org/10.1021/acs.iecr.5b04276.
Du, L., Chu, L. 2012. Understanding Anomalous Phase Behavior in Unconventional Oil
Reservoirs. Paper presented at the SPE Canadian Unconventional Resources
Conference, Calgary, Alberta, Canada. Society of Petroleum Engineers. DOI: http://
dx.doi.org/10.2118/161830-MS.
Honarpour, M.M., Nagarajan, N.R., Orangi, A. et al. 2012. Characterization of Critical
Fluid Pvt, Rock, and Rock-Fluid Properties - Impact on Reservoir Performance of
Liquid Rich Shales. Paper presented at the SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, USA. Society of Petroleum Engineers. DOI: http://
dx.doi.org/10.2118/158042-MS.
Kozeny, J., 1927. Ueber kapillare leitung des wassers im boden. Sitzungsber Akad 136
(2a), 271–306.
Lohrenz, J., Bray, B.G., Clark, C.R., 1964. Calculating viscosities of reservoir fluids from
their compositions. J. Petrol. Technol. 16 (10). http://dx.doi.org/10.2118/915-PA.
Michelsen, M.L., 1982a. The isothermal flash problem. Part I. stability. Fluid Phase
Equilibria 9 (1), 1–19. http://dx.doi.org/10.1016/0378-3812(82)85001-2.
Michelsen, M.L., 1982b. The Isothermal Flash Problem. Part Ii. Phase-Split Calculation.
Fluid Phase Equilibria 9 (1), 21–40. http://dx.doi.org/10.1016/0378-3812(82)
85002-4.
Nelson, P.H., 1994. Permeability-porosity relationships in sedimentary rocks. Log
Analyst 35 (3), 38–62.
Nojabaei, B., 2015. Phase Behavior and Flow Analysis of Shale Reservoirs Using a
Compositionally-Extended Black-Oil Approach, PhD Dissertation. The Pennsylvania
State University.
Nojabaei, B., Johns, R.T., Chu, L., 2013. Effect of capillary pressure on phase behavior in
tight rocks and shales. SPE Reserv. Eval. Eng. 16 (03), 281–289. http://dx.doi.org/
10.2118/159258-PA.
Nojabaei, B., Siripatrachai, N., Johns, R.T. et al. 2014. Effect of Saturation Dependent
Capillary Pressure on Production in Tight Rocks and Shales: A Compositionally-
Extended Black Oil Formulation. Paper presented at the SPE Eastern Regional
Meeting, Charleston, WV, USA. Society of Petroleum Engineers. DOI: http://dx.doi.
682
Journal of Petroleum Science and Engineering 152 (2017) 675–682
org/10.2118/171028-MS.
Pang, J., Zuo, J.Y., Zhang, D. et al. 2012. Impact of Porous Media on Saturation
Pressures of Gas and Oil in Tight Reservoirs. Paper presented at the SPE Canadian
Unconventional Resources Conference, Calgary, Alberta, Canada. Society of
Petroleum Engineers. DOI: http://dx.doi.org/10.2118/161143-MS.
Pederson, K.S., Christensen, P.L., 2007. Phase Behavior of Petroleum Reservoir Fluids.
CRC Press Taylor & Francis Group, Original edition. ISBN 978-0-8247-0694-4.
Rachford, H.H., Rice, J.D., 1952. Procedure for use of electrical digital computers in
calculating flash vaporization hydrocarbon equilibrium. J. Petrol. Technol. 4 (10).
http://dx.doi.org/10.2118/952327-G.
Rezaveisi, M., Sepehrnoori, K., Pope, G.A. et al. 2015. Compositional Simulation
Including Effect of Capillary Pressure on Phase Behavior. Paper presented at the SPE
Annual Technical Conference and Exhibition, Houston, Texas, USA. Society of
Petroleum Engineers. DOI: http://dx.doi.org/10.2118/175135-MS.
Sandoval, D., Yan, W., Michelsen, M.L. et al. 2015. Phase Envelope Calculations for
Reservoir Fluids in the Presence of Capillary Pressure. Paper presented at the SPE
Annual Technical Conference and Exhibition, Houston, Texas, USA. Society of
Petroleum Engineers. DOI: http://dx.doi.org/10.2118/175110-MS.
Sigmund, P.M., Dranchuk, P.M., Morrow, N.R., et al., 1973. Retrograde condensation in
porous media. Soc. Petrol. Eng. J. 13 (02), 93–104. http://dx.doi.org/10.2118/
3476-PA.
Tang, D.E. and Zick, A.A. 1993. A New Limited Compositional Reservoir Simulator.
Paper presented at the SPE Symposium on Reservoir Simulation, New Orleans,
Louisiana. Society of Petroleum Engineers. DOI: http://dx.doi.org/10.2118/25255-
MS.
Wang, D., Butler, R., Zhang, J., et al., 2012. Wettability survey in bakken shale with
surfactant-formulation imbibition. SPE Reserv. Eval. Eng. 15 (06), 695–705. http://
dx.doi.org/10.2118/153853-PA.
Wang, F.P. and Reed, R.M. 2009. Pore Networks and Fluid Flow in Gas Shales. Paper
presented at the SPE Annual Technical Conference and Exhibition, New Orleans,
Louisiana. Society of Petroleum Engineers. DOI: http://dx.doi.org/10.2118/124253-
MS.
Wang, S., Feng, Q., Javadpour, F., et al., 2016b. Breakdown of fast mass transport of
methane through calcite nanopores. J. Phys. Chem. C 120 (26), 14260–14269.
http://dx.doi.org/10.1021/acs.jpcc.6b05511.
Wang, S., Javadpour, F., Feng, Q., 2016a. Confinement correction to mercury intrusion
capillary pressure of shale nanopores. Sci. Rep. 6, 20160. http://dx.doi.org/10.1038/
srep20160.
Wang, S., Javadpour, F., Feng, Q., 2016d. Molecular dynamics simulations of oil
transport through inorganic nanopores in shale. Fuel 171, 74–86. http://dx.doi.org/
10.1016/j.fuel.2015.12.071.
Wang, S., Javadpour, F., Feng, Q., 2016e. Fast mass transport of oil and supercritical
carbon dioxide through organic nanopores in shale. Fuel 181, 741–758. http://
dx.doi.org/10.1016/j.fuel.2016.05.057.
Wong, T.W., Firoozabadi, A., Aziz, K., 1990. Relationship of the volume-balance method
of compositional simulation to the Newton-Raphson method. SPE Reserv. Eng. 5
(03), 415–422. http://dx.doi.org/10.2118/18424-PA.
Wu, K., Chen, Z., Li, X., et al., 2016. Methane storage in nanoporous material at
supercritical temperature over a wide range of pressures. Sci. Rep. 6, 33461. http://
dx.doi.org/10.1038/srep33461.
Young, L.C., Stephenson, R.E., 1983. A generalized compositional approach for reservoir
simulation. Soc. Petrol. Eng. J. 23 (05), 727–742. http://dx.doi.org/10.2118/10516-
PA.