Journal of Petroleum Science and Engineering 166 (2018) 891–899

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

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

Study on the rock-electric and the relative permeability characteristics in T

porous rocks based on the curved cylinder-sphere model
He Meng
State Key Laboratory of Marine Geology, Tongji University, Shanghai, 200092, China

A R T I C LE I N FO A B S T R A C T

Keywords: Usually, in conventional reservoirs, the rock-electric properties can be explained by Archie's law, and the relative
Pore structure permeability is only considered as a function of fluid saturation. However, as for the complex reservoirs, Archie's
Fluid distribution law fails to accurately describe the rock-electric properties and the relative permeability is not only affected by
Wettability saturation, the reason is that both rock-electric and relative permeability characteristics are influenced by multi-
Curved cylinder-sphere model
factors, including the complex pore structure, fluid distribution, and wettability. Currently, there lacks the
Rock-electric characteristics
Relative permeability
theoretical method to comprehensively study the multi-factors effect on rock-electric and relative permeability
characteristics. In this paper, the curved cylinder-sphere model is developed to illustrate the complex pore-throat
structure, fluid distribution, and wettability in real rocks, which can be characterized by two important para-
meters: the ratio Cd of the curved cylinder radius to the sphere radius and the tortuosity τ. Based on the curved
cylinder-sphere model, and by using Ohm's law, one can carry out the research on the effect of pore geometry,
fluid distribution, and wettability on rock-electric characteristics. Moreover, by combining the curved cylinder-
sphere model with Li's model, the effect of pore structure, fluid distribution, and wettability can be incorporated
into the relative permeability calculation model. By means of the numerical simulation and analysis, it comes to
conclusions that pore structure, fluid distribution, and wettability are the principal factors affecting both rock-
electric and relative permeability characteristics, which may cause different formation factor, resistivity index,
and relative permeability even in those reservoirs with the same porosity and fluid saturation. Besides, the study
demonstrates that the complex pore structure, fluid distribution, and wettability may cause the non-Archie
phenomenon of rock-electric characteristics, in addition, the complex pore structure can lead to the decrease of
water relative permeability and the increase of oil relative permeability, and when the rock wettability changes
from water-wet to oil-wet, water relative permeability increases and oil relative permeability decreases.
Furthermore, by the comparison of the simulation results and Lab data, the important effect of complex pore
structure and fluid saturation is confirmed.

1. Introduction 2007) have extensively studied the non-Archie phenomenon in porous

rocks.
In terms of the conductive model in two-phase flow medium, the Besides, some studies indicate that the relative permeability can
classic Archie equation can give a good description of rock-electric intuitively reflect oil-water seepage characteristics, which is related to
characteristics in pure sandstone reservoirs with simple pore structure. rock-electric characteristics, and the study of oil-water seepage char-
However, the study shows that there always exists non-Archie phe- acteristics is the key to water flooding development of low permeability
nomenon and low-resistivity oil layers in complex reservoirs caused by reservoir. Generally, the relative permeability can be measured in la-
some factors, such as the complex pore structure, fluid distribution, and boratory by using the steady state and unsteady state methods.
wettability, which cannot be well explained by Archie law. As the water However, as for those rocks with the low permeability and complex
saturation decreases and the pore structure becomes complex, the non- pore structure, it is not only time-consuming and expensive but also
Archie phenomenon becomes more evident, and non-Archie behaviors hard to obtain directly by the lab measurement. Later, some studies
of porous rocks (i.e., F − ϕ and I − Sw relationships are not linear on a published have focused on the estimation of the relative permeability
log-log scale) have been increasingly observed in electrical logging in- from capillary pressure data. Purcell (1949) established the relationship
terpretation. Diederix (1982), Li (1989), Worthington and Pallatt between the relative permeability and pore size, capillary pressure.
(1992), Jing et al. (1993), Montaron (2009), and Padhy et al. (2006, According to Gates and Leitz (1950) report, this relationship has been

E-mail address: [email protected].

https://doi.org/10.1016/j.petrol.2018.03.085
Received 17 November 2017; Received in revised form 21 March 2018; Accepted 23 March 2018
Available online 30 March 2018
0920-4105/ © 2018 Elsevier B.V. All rights reserved.
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

applied to multiphase flow in porous media. Burdine (1953) then took saturation and at a specific water saturation of Sw can be derived:
the tortuosity factor into consideration in the model, which is a function
τ 2o
of fluid saturation. By summarizing the previous work, Corey (1954) Ro = Rw
ϕ (1)
and Brooks and Corey (1966) modified the capillary pressure as a
power function of wetting-phase saturation, and proposed a new model
τ t2
to determine the relative permeability, which has been widely used in Rt = Rw
ϕSw (2)
the field. However, capillary pressure is not only hard to measure but
also not practical for field data processing. Li (2005, 2008, 2011) found where τo is the tortuosity at a 100% water saturation, τo = L′/ L , a di-
a relationship between relative permeability and resistivity index from mensionless quantity whose value is always ≥ 1; τt is the tortuosity at a
the analogy between fluid flow and electric flow properties. Pairoys specific water saturation of Sw, τt = l′/ L , a dimensionless quantity
et al. (2013) improved Li's model by considering residual non-wetting whose value is always ≥ 1; Ro and Rt are the resistance of rocks at a
phase saturation. Ma et al. (2015) derived two-phase relative perme- saturation of 100% and a specific saturation Sw, respectively; Rw is the
ability from resistivity by combining Poiseuille's law with Darcy's law resistance of brine; ϕ is rock porosity.
and introducing tortuosity ratio into the model. Ge (2015) predicted the According to Archie's law, the resistivity index can be written as:
relative permeability based on theory of coupled electricity-seepage
τ t2 1
and capillary bundle model. However, few work is conducted to study I=
and model the important effect of pore structure, fluid distribution, and τ 2o Sw (3)
wettability on the relative permeability. For rocks with simple pore structure, both τo and τt take the value of
In this work, the objective of the study is to study the effect of pore 1, then the resistivity index becomes:
structure, fluid distribution, and wettability on the rock-electric and
1
relative permeability characteristics in two-phase flow medium (water I=
is wetting phase, and oil is non-wetting phase). A curved cylinder- Sw (4)
sphere model is proposed to describe the complex pore-throat structure, Eq. (4) is considered as the simplest form of Archie's law, which can
fluid distribution, and wettability properties, which is described by two be used to describe the resistivity index property of the simple capillary
parameters: the tortuosity and the ratio of curved cylinder radius to model, while it can't illustrate the complex pore structure.
sphere radius. By the comparison of the simulation results and the ex- Li (2008) found a relationship between relative permeability and
perimental data, it shows that the formation factor and the resistivity resistivity index from the analogy between fluid flow and electric flow
index are greatly affected by pore structure, fluid distribution, and properties. Li's model is expressed as:
wettability, and the theoretical results are in good agreement with ex-
1
perimental data, besides, by considering the complex pore structure, Krw = Sw∗
I (5)
fluid distribution, and wettability, the non-Archie phenomenon can be
interpreted. Furthermore, by combining the curved cylinder-sphere Sw − Swi
Sw∗ =
model with Li's model, the pore structure parameters can be introduced 1 − Swi (6)
into the relative permeability model, and the modified model can be
used to model and explain the effect of the complex pore-throat where Sw∗
is normalized saturation, dimensionless unit, Sw is the wet-
structure, fluid distribution, and wettability on the relative perme- ting phase saturation, dimensionless unit, and Swi is the irreducible
ability. saturation of the wetting phase, dimensionless unit.
Substituting Eq. (3) into Eq. (5), the relative permeability is ex-
pressed as:
2. Resistivity index and relative permeability in rocks with simple
pore structure τ 2o ∗
Krw = Sw S w
τ t2 (7)
In real reservoirs, pore structure is very complex. However, for the
sake of simplicity, pore model should be simplified to illustrate the pore Burdine (1953) gave an empirical expression of tortuosity ratio,
structure. which is a function of wetting phase saturation:
Usually, reservoir rock is simplified into a unit volume cylinder, and τo S − Swi
= w = Sw∗
pore structure is simplified as a curved cylindrical pore containing oil- τt 1 − Swi (8)
water two-phase flow, as shown in Fig. 1, A is the apparent cross-sec-
Substituting Eq. (8) into Eq. (7) yields the relative permeability:
tion area of rock, and is equal to 1.0, L is the length of rock, and equals
1.0; Aw is the apparent cross-section area of effective water flow Krw = Sw∗3 Sw (9)
pathway, l′ is the length of the tortuous water flow pathway and L′ is
the length of the tortuous cylindrical pore. Based on Archie's law Note that the derived Eq. (9) is similar to the semi-empirical relative
(Archie, 1942; Cai et al., 2017), the resistance of rock at a 100% water permeability model proposed by Corey (1954), and the relative per-
meability can be easily inferred from water saturation, but it only
contains the effect of fluid saturation, which is not suitable for complex
reservoir rocks.

3. Study on rock-electric characteristics of complex rocks based

on the curved cylinder-sphere model

3.1. Rock-electric characteristics in single-phase flow medium

Here a real reservoir rock is still simplified into a unit volume cy-
linder with a cross-section area of 1.0 and a length of 1.0, however, the
complex pore structure is considered as a curved cylinder-sphere
model, which is composed of sphere pore and curved cylinder pore, and
Fig. 1. The oil-bearing curved cylinder model. the pore space is fully saturated with water, as shown in Fig. 2. The pore

892
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 2. The curved cylinder-sphere model.

structure is described by two parameters: the tortuosity and the curved Fig. 3. Comparison of lab data and simulation results on formation factors, the
cylinder-sphere radius ratio. The tortuosity and the curved cylinder- tortuosity τ = 1.0.
sphere radius ratio Cd are defined as (Johnson and Schwartz, 1989;
Epstein, 1989; Shang et al., 2003; Liu and Tang, 2013):
L′
τ=
L (10)

Rc
Cd =
Rs (11)
where L′ is the effective length of the curved cylinder, L is the straight-
1
line length of the curved cylinder, L = 2 − Rs2 − Rc2 , Rc is the curved
cylinder radius (R c ≤ Rs ), Rs is the sphere radius ( πRs2 < 1 , Rs < 1/ π
), and it is easily found out that 0 < Cd ≤ 1.
As shown in Fig. 2, the volume V of the curved cylinder-sphere
model is divided into two parts: the curved cylinder volume Vc and the
sphere volume Vs without spherical crowns, thus, the pore space vo-
lume V is the volume sum of two parts. Due to the unit volume rock,
porosity ϕ is equal to the pore volume V:
Fig. 4. Comparison of lab data and simulation results on formation factors, the
Rs2 − Rc2 tortuosity τ = 1.25.
ϕ = V = Vs + Vc = 2 ∫ Ss (r ) dr + 2Sc L′
0 (12)
where Ss (r ) and Sc are the cross-section area of the sphere and the
curved cylinder, respectively, Ss (r ) = π (Rs2 − r 2) , Sc = πRc2 . Thus, the
porosity ϕ can be solved as:
2π 3 1
ϕ= Rs 1 − Cd2 (2 + Cd2) + 2πτRs2 Cd2 ⎛ − Rs 1 − Cd2 ⎞
3 ⎝2 ⎠ (13)
Similarly, the resistance of reservoir rock Ro can be expressed as the
sum of the resistance of the curved cylinder part and sphere part.
According to Ohm's law and Archie law, the formation factor F can be
obtained by:
Rs2 − Rc2
R L′ 2dr Lτ 2dr
F= o =
Rw Sc
+ ∫ Ss (r )
=
πRc2
+ ∫ π (Rs2 − r 2)
0

(1 − 2Rs 1 − Cd2 ) τ 1 1+ 1 − Cd2

= + ln Fig. 5. Comparison of lab data and simulation results on formation factors,
πRs2 Cd2 πRs 1− 1 − Cd2 (14) Cd = 0.25.

As indicated by Wei et al. (2015), the pore radius is an important

factor for the formation factor, Eq. (14) also shows the relationship follows:
between the formation factor and the pore radius, when porosity, the As stated above, the variations of τ and Cd indicate the change of the
tortuosity, and the ratio of curved cylinder radius to sphere radius are pore structure. For the same porosity, the formation factor F still
given, the values of curved cylinder radius and sphere radius can be changes with τ and Cd , which means that rock-electric characteristics
obtained by using Eq. (13). are greatly affected by the pore structure.
Fig. 2 illustrates the complex pore structure, and different pore As seen from Fig. 3, when the value of Cd becomes small (Cd = 0.25),
structure could be regarded as the variations of τ and Cd . Using Eqs. the pore structure of the rock tends to be complex, and the relationship
(13) and (14), when τ and Cd take different values, the relationship between porosity ϕ and formation factor F presents the nonlinear
between formation factor and porosity for rocks with different pore change; when the value of Cd becomes large (Cd = 0.95), the pore
structure could be obtained, as shown in Figs. 3–5. The simulation re- structure of the rock tends to be simple, and the F − ϕ relationship is
sults in all these Figures illustrate some observations summarized as close to linear change, which is consistent with the rock-electric

893
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 6. The oil-bearing curved cylinder-sphere model.

Fig. 7. Comparison of simulation results and lab data on the resistivity index I,
characteristics of classic Archie law. By the comparison of the simula-
the tortuosity τ = 1.0.
tion results and Lab data, it comes to the conclusion that the simulation
results are in good agreement with lab data by considering the complex
rs2− rc2
pore structure. (1 − 2Rs 1 − Cd2 ) τ 2dr
In addition, the increase of tortuosity also leads to more complex Ft =
π (Rc2 − rc2)
+ ∫ π (Rs2 − r 2) − π (rs2 − r 2)
0
pore structure, Figs. 4 and 5 show that the increasing tortuosity in-
Rs2 − Rc2
creases formation factor F and the nonlinear degree of the F − ϕ re- 2dr (1 − 2Rs 1 − Cd2 ) τ
lationship, it proves that the complex pore structure is critical for the + ∫ π (Rs2 2
− r 2) − πrc
=
π (Rs2 Cd2 − rs2 Cdo
2
)
study of rock-electric properties, which can cause the non-Archie phe- rs2− rc2

nomenon. Rs2 − rs2 Cdo

2 − r 1 − C2
⎛ s do ⎞
2
ln
2rs 1 − Cdo 1 ⎜ Rs2 − rs2 Cdo
2 + r 1 − C2
s do
⎟
+ + ⎜ ⎟
π (Rs2 − rs2) π Rs2 − rs2 Cdo
2 Rs2 − rs2 Cdo
2 −R 1 − Cd2
3.2. Rock-electric characteristics in two-phase flow medium ⎜⎜− ln
s
⎟
Rs2 − rs2 Cdo
2 +R
s 1 − Cd2 ⎟
⎝ ⎠
When oil flows into the reservoir rock, the shape of oil is assumed to (18)
be the same as pore structure in the curved cylinder-sphere model, as
shown in Fig. 6. As described above, a reservoir rock is thought as a unit Then, by combining Eqs. (14) and (18), the resistivity index can be
volume cylinder. In Fig. 6, the pore space is filled with oil-water two- determined by:
phase flow. The structure parameters of oil bead are defined as the Ft
follows: I=
F (19)

l′ As for oil-bearing curved cylinder-sphere model, Cd and τ can be

τo = used to describe the pore structure, Cdo and τ can be used to illustrate
l (15)
the oil and water distribution. Fig. 7 shows that the resistivity index I
rc changes with different combinations between Cd and Cdo. When Cdo is
Cdo = equal to Cd, it means that the distribution of oil bead in pore space is in
rs (16)
accordance with the pore structure, and I − Sw curve displays the linear
where τo is the tortuosity of oil bead, in this work, for the sake of relationship, which is identical to the resistivity index characteristic of
simplicity, the oil bead tortuosity τo is thought to be the same as the classic Archie law; when Cdo is greater than Cd, the rock resistivity index
pore tortuosity τ , namely τo = τ . l′ is the effective length of the curved increases sharply and presents the nonlinear relationship of I − Sw ;
cylinder of oil bead, l is the straight-line length of the curved cylinder particularly, compared with other combinations of Cdo and Cd , when Cdo
1 is smaller than Cd, the resistivity index in low water saturation becomes
of oil bead, l = 2 − rs2 − rc2 , Cdo is the curved cylinder-sphere radius
ratio of oil bead, rc is the curved cylinder radius of oil bead (rc ≤ R c ), rs is much lower, which can be considered as the low resistivity character-
istic in the oil-bearing zone. Therefore, different combinations between
the sphere radius of oil bead (rs ≤ Rs ).
Cd and Cdo, which reflect different fluid distribution characteristics, may
According to Ohm's law and Archie law, the generalized formation
result in the high resistivity or the low resistivity in the oil-bearing
factor Ft can be obtained by:
zone. Besides, the simulation results show that due to the uneven dis-
rs2− rc2 Rs2 − Rc2
tribution of oil (Cdo = 0.50, Cd = 0.45), I − Sw has a nonlinear re-
Rt 2dr 2dr lationship, which is confirmed by Lab data in Fig. 7. It is easily found
Ft =
Rw
= ∫ Ss − Sso
+ ∫ Ss − Sco that rock-electric characteristics are not only related to the pore
0 rs2− rc2
1
structure, but also related to the fluid distribution in pore space.
2
2dr −b ± b2 − 4ac In the paper, Fig. 8 shows that the increasing tortuosity decreases
+ ∫ Sc − Sco 2a the resistivity index. In order to explain the trend in greater detail, the
Rs2 − Rc2 (17) tortuosity effect on the electric characteristics of two parts (curved
cylinder and sphere parts) is thoroughly discussed and analyzed.
where Rt is the resistance of two-phase flow rock, Sso (r ) and Sco are the According to the resistance law and Archie law, the resistivity index
cross-section area of the sphere and the curved cylinder of oil bead, I can be written as:
respectively, Sso (r ) = π (rs2 − r 2) , Sco = πrc2 . The above Eq. (17) can be
simplified as: (Rso + R co)/ Rw F
I= = t
(Rsw + R cw )/ Rw F (20)

where Rso is the resistance of partially water-saturated sphere part, R co

894
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 10. The resistance changes of sphere part with the tortuosity τ.
Fig. 8. Comparison of simulation results and lab data on the resistivity index I,
ϕ = 0.25, Cd = 0.45, Cdo = 0.50.

is the resistance of partially water-saturated curved-cylinder part, Rsw is

the resistance of fully water-saturated sphere part, R cw is the resistance
of fully water-saturated curved-cylinder part, Rw is the resistance of
brine, Ft is the generalized formation factor, F is the Archie formation
factor.
First, both porosity and saturation are held constant,
ϕ = 0.25, Sw = 0.30 , then the change of tortuosity affects the pore
structure and fluid distribution, which causes the changes of the rock
resistance. Figs. 9 and 10 illustrate the changes of R co , R cw , Rso , and Rsw
with the tortuosity, respectively, it shows that the difference between
R co and R cw decreases with the increase of tortuosity, and the difference
between Rso and Rsw also decreases with the increase of tortuosity, thus,
based on the above Eq. (20), it is easily known that the difference be-
tween the partially water-saturated rock resistance (Rso + R co ) and the
fully water-saturated rock resistance (Rsw + R cw ) decreases with the
Fig. 11. Rock-electric parameters at different tortuosity τ, ϕ = 0.25, Sw = 0.30,
increase of tortuosity, it also means the difference between Ft and F
Cd = 0.45, and Cdo = 0.50.
decreases with the increase of tortuosity, which eventually causes the
decrease of the resistivity index I . The analysis results are consistent
with the simulation results, as shown in Fig. 11. Krw
⎛ ⎛1 − 2Rs 1 − Cd2 ⎞ τ
⎜ ⎟
1 + 1 − Cd2
⎞
1
⎜⎝ ⎠
+ ln ⎟ Sw∗
4. The effect of pore structure on relative permeability based on πRs2 Cd2 πR s 1− 1 − Cd2
⎜ ⎟
the curved cylinder-sphere model = ⎝ ⎠
⎛ Rs2 − rs2 Cdo
2 − r 1 − C2
s do ⎞
⎛1 − 2R 1 − C 2 ⎞ τ ln
Substituting Eq. (19) into Eq. (5), one can obtain the relative per- ⎜ s d⎟ 2
2rs 1 − Cdo 1
⎜ Rs2 − rs2 Cdo
2 + r 1 − C2
s do
⎟
⎝ ⎠
+ +
meability of wetting phase, which contains the pore structure effect: π (Rs2 Cd2 − rs2 Cdo
2 ) π (Rs2 − rs2) 2 ⎜
Rs2 − rs2 Cdo
⎟
π Rs2 − rs2 Cdo
2 −R
s 1 − Cd2
⎜⎜− ln ⎟
Rs2 − rs2 Cdo
2 +R
s 1 − Cd2 ⎟
⎝ ⎠
(21)
According to Li and Horne (2002, 2006), the relative permeability
of the wetting phase can be determined by Purcell model, and the re-
lative permeability of the non-wetting phase can be better described by
Brooks-Corey model. By combining Purcell model and Brooks-Corey
model, the relative permeability relationship between wetting phase
and non-wetting phase can be expressed as:

Kro = (1 − Sw∗ )2 (1 − Krw ) (22)

where Kro is the relative permeability of non-wetting phase, di-

mensionless unit.
Substituting Eq. (21) into Eq. (22), one can obtain the relative
permeability of non-wetting phase including the pore structure effect.
As shown in above Eqs. (21) and (22), the relative permeability is
related to both pore structure and fluid distribution, and the combi-
nations of curved cylinder radius rc and sphere radius rs directly de-
termines oil saturation and distribution properties in pore space. In the
Fig. 9. The resistance changes of cylinder part with the tortuosity τ. first instance, in order to study the effect of fluid distribution on relative

895
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 12. The changes of relative permeability under different pore structure
parameter Cd and tortuosity τ, Cd = 0.50.

Fig. 14. Spectrogram of Cdo, Cd = 0.50.

permeability, the pore structure is fixed, namely both ϕ and Cd remain
constant, then pore cylinder radius R c and sphere radius Rs can be
determined, meanwhile, oil cylinder radius rc also keeps constant
(rc < R c ), oil sphere radius rs can range from rc to Rs (rc < rs < Rs ), and
oil cylinder-sphere radius ratio Cdo ranges from rc / Rs to 1, which reflects
the change of fluid distribution.
As seen in Fig. 10, the relative permeability of water phase increases
with the increasing Cdo , and the relative permeability of oil phase de-
creases with the increasing Cdo , which indicates that the more uneven
the fluid distribution is (the smaller Cdo becomes), the smaller the re-
lative permeability of water phase is and the larger the relative per-
meability of oil phase is; besides, Fig. 12 shows that the complex tor-
tuosity also affects the relative permeability, both water relative
permeability and oil relative permeability decrease with the increasing
tortuosity τ .
Next, in order to study the effect of pore structure on the relative
permeability, Fig. 13 describes how the relative permeability changes
with the pore structure parameter Cd , when Cd ranges from 0.5 to 0.2,
the pore structure becomes more complex, the relative permeability of
water phase decreases and the relative permeability of oil phase in-
Fig. 15. Spectrogram of normalized saturation, Cd = 0.50.
creases, which demonstrates that the complex pore structure is the
important factor influencing the relative permeability.
Furthermore, in order to study the effect of fluid distribution and cylinder–sphere radius ratio Cdo of oil bead at different combination
saturation on the relative permeability in complete detail, the pore between rc and rs , which describes the various fluid distribution; Fig. 15
structure parameter Cd and porosity ϕ still remain constant, RS and RC gives the change of water saturation with the simultaneous change of rc
can be calculated by Eq. (13) (RS = 0.4, RC = 0.2), which determine the and rs , which illustrates the process from water saturation to oil sa-
ranges of rS and rC. Then the fluid distribution and saturation conditions turation; furthermore, the spectrograms of water and oil relative per-
can be modeled by changing and choosing different combinations of rc meability are depicted in Figs. 16 and 17, respectively, by the
and rs , as shown in Figs. 14 and 15; Fig. 14 shows the spectrogram of

1
Relative Permeability, fraction

0.8

0.6 Krw from new model, Cd=0.50

Kro from new model, Cd=0.50
Krw from new model, Cd=0.35
0.4 Kro from new model, Cd=0.35
Krw from new model, Cd=0.20
Kro from new model, Cd=0.20
0.2

0
0.1 0.3 0.5 0.7 0.9 1
Cylinder-Sphere Radius Ratio Cdo of Oil, fraction

Fig. 13. The changes of relative permeability under different pore structure
parameter Cd and oil bead structure parameter Cdo, τ = 1.0. Fig. 16. Spectrogram of water relative permeability Krw, Cd = 0.50.

896
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 19. Spectrogram of oil relative permeability Kro, Cd = 0.25.

Fig. 17. Spectrogram of oil relative permeability Kro, Cd = 0.50.

The study shows that there is a water film or oil film with a thick-
comparison of Figs. 14–17, one can clearly understand the effect of both
ness of h on the pore surface of water-wet or oil-wet rock, as shown in
the fluid saturation and the fluid distribution on the relative perme-
Fig. 6. Hirasaki (1991) pointed out that the wetting degree is closely
ability. By the comparison among these figures, it is easily found that
related to the thickness of water or oil film in the oil-water-rock system.
water relative permeability increases and oil relative permeability de-
Under the water-wet condition, suppose that there exists a thin layer
creases with the increase of Cdo and water saturation Sw , even though
of water film on the pore surface, which provides the conductive
Cdo is small, water relative permeability is still high at high water sa-
channel and increases the irreducible water saturation, the average
turation.
thickness of water film is denoted as h . It is easily found that the gen-
Also, when porosity is constant and Cd decreases from 0.5 to 0.25,
eralized formation factor Ft is identical to Eq. (17), however, the range
the range of cylinder pore radius rc decreases by nearly half, while the
of rc and rs change, rc ranges from 0 to R c − h , and rs ranges from 0 to
range of sphere pore radius rs is almost unchanged, as shown in Figs. 18
Rs − h .
and 19. Moreover, by the comparison between Figs. 16 and 18, when
Similarly, under the oil-wet condition, there exists a thin layer of oil
both water saturation and oil bead structure remain constant, such as
film on the pore surface, which increases the residual oil saturation and
rs = 0.2, rc = 0.1, it is obvious that water relative permeability in Fig. 18
decreases the effective cross-sectional area of water path, the average
is less than that in Fig. 16, thus, due to the pore structure difference, the
thickness of water film is also denoted as h . Besides, R c is replaced by
relative permeability may be different in those reservoirs with the same
R c − h , and Rs is replaced by Rs − h . The generalized formation factor
porosity and fluid saturation.
Ft can be rewritten as:
rs2− rc2 (R s − h)2 − (R c − h)2
5. The effect of wettability on rock-electric and relative Rt 2dr 2dr
permeability characteristics based on the curved cylinder-sphere
Ft =
Rw
= ∫ Ss − Sso
+ ∫ Ss − Sco
0 rs2− rc2
model 1
2
2dr
In fact, besides the pore structure and fluid distribution, the wett- + ∫ Sc − Sco
ability is also related to the non-Archie phenomenon and relative per- (R s − h)2 − (R c − h)2 (23)
meability results. In the following research, the effect of wettability on Using the same method stated above, by combining with Li's model,
rock-electric and relative permeability characteristics is discussed and the effect of wettability on the relative permeability also can be studied.
analyzed. In order to model the rock-electric and relative permeability char-
acteristics in different wetting conditions, the average thicknesses of
water film and oil film are set to 0.0041 and 0.0205, respectively. The
rock-electric and relative permeability characteristics in water-wet and
oil-wet conditions are depicted in Figs. 20 and 21.
Many scholars have conducted a lot of experimental research and
theoretical analysis on the influence of wettability on the conductivity
of rocks. In oil wet rocks, with the decrease of water saturation, re-
sistivity index increased; in water wet rocks, a layer of water film forms
on the surface of pore, which provides an additional current channel
and increases the content of bound water. Therefore, under the same
water saturation, the resistivity index of oil wet rocks is significantly
higher than that of water wet rocks, as shown in Fig. 20, besides, Fig. 20
also shows that the change of wettability leads to the nonlinear change
of the I − Sw relationship, which means that the rock wettability also
can cause the non-Archie phenomenon; furthermore, the wettability
will also affect the relative permeability characteristics of oil-water
two-phase flow, when the rock wettability changes from oil wet to
water wet, oil relative permeability tends to increase, water relative
Fig. 18. Spectrogram of water relative permeability Krw, Cd = 0.25. permeability tends to decrease, and the isoperm point in the relative

897
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

Fig. 20. The changes of resistivity index in water-wet and oil-wet conditions. Fig. 22. Relative permeability calculated from Li model and new model. in
Boise sandstone ( ϕ = 0.32) at a temperature of 300 °F.

Fig. 21. The changes of relative permeability in water-wet and oil-wet condi-
tions.
Fig. 23. Relative permeability calculated from Li model and new model. in
Berea sandstone ( ϕ = 0.204) at a temperature of 300 °F.
permeability curve moves to the right, these characteristics are shown
in Fig. 21. Generally, the rock-electric and relative permeability char-
acteristics in water-wet and oil-wet conditions simulated by this model
are consistent with previous experimental results.

6. Comparison and validation

The experimental data of resistivity measured by Sanyal (1972) in

different rocks (Boise sandstone, Berea sandstone, and limestone) were
used to test the effect of pore structure and fluid distribution considered
in this work. Firstly, the values of oil/water relative permeability were
calculated using Li's model. Secondly, the oil/water relative perme-
ability data were simulated using the curved cylinder-sphere model.
Finally the results of relative permeability inferred from two models
were compared.
Fig. 22 shows that oil and water relative permeability in the Boise
sandstone core sample calculated from Li model is close to the simu-
Fig. 24. Relative permeability calculated from Li model and new model. in
lation results in the condition of the simple pore structure
limestone ( ϕ = 0.19) at a temperature of 300 °F.
(Cd = Cdo = 1.0 ). Fig. 23 gives oil and water relative permeability data
in the Berea sandstone sample, which were calculated from Li model
and were modeled by the new model, respectively. The results are water relative permeability are affected by the pore structure, water
compared and shown in Fig. 23, when both Cd and Cdo are equal to 0.5, relative permeability (wetting phase) decreases and oil relative per-
the relative permeability data inferred from Li model are almost equal meability (non-wetting phase) increases with more complex pore
to those simulated using the curved cylinder-sphere model. As for the structure and more uneven fluid distribution (Cd and Cdo become
limestone core sample, the simulation results and the relative perme- smaller). By taking pore sructure and fluid distribution into account,
ability inferred from Li model are depicted in Fig. 24. One can see that the simulation results are well matched with the experimental data,
the oil and water relative permeability inferred from Li model are al- which demonstrates that the relative permeability can be influenced by
most equal to the simulation results when both Cd and Cdo are 0.6. pore structure and fluid distribution as well as fluid saturation.
Furthermore, one can see from Figs. 22–24 that both the oil and

898
H. Meng Journal of Petroleum Science and Engineering 166 (2018) 891–899

7. Conclusions porous media: a review. Earth Sci. Rev. 171, 419–433.

Corey, A.T., 1954. The interrelation between gas and oil relative permeabilities. Prod.
Mon. 19 (1), 38–41.
In this paper, the curved cylinder-sphere model is proposed to il- Diederix, K.M., 1982. Anomalous relationships between resistivity index and water sa-
lustrate the complex pore structure, fluid distribution, and wettability turations in the Rotliegend Sandstone. In: SPWLA 23th Annual Logging Symposium,
of rocks, in fact, the curved cylinder part and sphere part correspond to Corpus Christi, Texas, pp. 6–9.
Epstein, N., 1989. On tortuosity and the tortuosity factor in flow and diffusion through
throat and pore in real rocks, besides, when there exists water film or porous media. Chem. Eng. Sci. 44, 777–779.
oil film on the pore surface, the rock is assumed to be water-wet or oil- Gates, J.I., Leitz, W.J., 1950. Relative permeabilities of California cores by the capillary
wet. Based on the curved cylinder-sphere model, one can study and pressure method. In: API Meeting, Los Angeles, California, pp. 286.
Ge, X., 2015. Predicting the relative permeability of water phase based on theory of
analyze rock-electric characteristics in fully and partly saturated rocks. coupled electricity-seepage and capillary bundle model. Open Pet. Eng. J. 8 (1),
The study shows that the complex pore structure, fluid distribution, and 344–349.
wettability are the important factors resulting in non-Archie phenom- Hirasaki, G.J., 1991. Wettability: fundamentals and surface forces. SPE Form. Eval. 6 (2),
217–226.
enon, which may cause the nonlinear change of the formation factor
Jing, X.D., Gillespie, A., Trewin, B.M., 1993. Resistivity index from non-equilibrium
and the resistivity index on a log-log scale, and the change of fluid measurements using detailed in-situ saturation monitoring. In: Offshore European
distribution may result in the complex conductive path, which may lead Conference, SPE-26798-Aberdeen, pp. 7–10.
to high resistivity or low resistivity in oil-bearing zones. Thus, due to Johnson, D.L., Schwartz, L.M., 1989. Unified theory of geometrical effects in transport
properties of porous media. In: SPWLA 30th Annual Logging Symposium, Soc. Of
the difference of pore structure and fluid distribution, there probably Prof. Well Log Anal., Houston, TX.
displays different rock-electric characteristics for those reservoirs with Li, K., Horne, R.N., 2002. Experimental verification of methods to calculate relative
the same porosity or fluid saturation. Moreover, by combining the permeability using capillary pressure data. In: Proceedings of the 2002 SPE Western
Region Meeting/AAPG Pacific Section Joint Meeting, Anchorage, Alaska.
curved cylinder-sphere model with Li's model, pore structure para- Li, K., 2005. A semi-analytical method to calculate relative permeability from resistivity
meters can be introduced into the relative permeability calculation well logs. In: SPE Annual Technical Conference and Exhibition, Dallas, USA.
model, which is beneficial to better understand the effect of pore Li, K., Horne, R.N., 2006. Comparison of methods to calculate relative permeability from
capillary pressure in consolidated water-wet porous media. Water Resour. Res. 42
structure, fluid distribution, and wettability on the seepage character- (6), 6405–6408.
istics. The simulation results prove that the relative permeability can be Li, K., 2008. A new method for calculating two-phase relative permeability from re-
greatly affected by pore structure, fluid distribution, and wettability, sistivity data in porous media. Trans. Porous Media 74 (1), 21–33.
Li, K., 2011. Interrelationship between resistivity index, capillary pressure and relative
when pore structure becomes more complex and fluid distribution be- permeability. Trans. Porous Media 88 (3), 385–398.
comes more uneven, water relative permeability is smaller and oil re- Li, N., 1989. Generalized resistivity-porosity and resistivity-oil saturation relationships.
lative permeability is larger, in addition, compared with the water-wet Chin. J. Geophys. 32 (5), 580.
Liu, T.Y., Tang, T.Z., 2013. Study of rock conductive mechanism based on pore structure.
condition, water relative permeability increases and oil relative per-
Chin. J. Geophys 56 (5), 674–684.
meability decreases in the oil-wet condition. Ma, D., Liu, C., Cheng, C., 2015. New relationship between resistivity index and relative
Therefore, the curved cylinder-sphere model not only theoretically permeability. J. Energy Resour. Tech. 137 (3), 032904.
proves the important effect of pore structure, fluid distribution, and Montaron, B., 2009. Connectivity theory – a new approach to modeling non-Archie rocks.
Petrophysics 50, 102–115.
wettability on formation factor, resistivity index, and relative perme- Padhy, G.S., Ioannidis, M.A., Lemaire, C., Coniglio, M., 2006. Measurement and inter-
ability but also provides an effective way to study rock-electric and pretation of non-archie resistivity behavior in model and reak vuggy carbonates. Soc.
seepage characteristics in complex reservoirs. Core Anal. 11, 1–11.
Padhy, G.S., Ioannidis, M.A., Lemaire, C., 2007. Interpretation of the non-Archie re-
sistivity behavior in Vuggy Carbonates: use of pore structure parameters. In: SPWLA
Acknowledgements Middle East Reg. Symp, pp. 1–14.
Pairoys, F., Al-Zoukani, A., Keskin, A., 2013. Correlation between resistivity and petro-
physical parameters of a carbonate rock during two-phase flow displacements. In:
This work is supported by National 863 Program of China SPE Saudi Arabia Section Technical Symposium and Exhibition. Society of Petroleum
(2006AA06Z214), National Natural Science Foundation of China Engineers.
(41476027), all research support funds are greatly appreciated. Purcell, W.R., 1949. Capillary pressures–their measurement using mercury and the cal-
culation of permeability. Trans. AIME 186, 39.
Sanyal, S.K., 1972. The Effect of Temperature on Electrical Resistivity and Capillary
References Pressure Behavior of Porous media. Ph.D. Report. Stanford University, Stanford,
California.
Shang, B.Z., Jeffry, Hamman, G., Donald, H., 2003. A Physical Model to Explain the First
Archie, G.E., 1942. The electrical resistivity log as an aid in determining some reservoir
Archie Relationship and beyond. SPE, pp. 84300.
characteristics. AIME Pet. Tech. 5 (2), 1–8.
Wei, W., Cai, J., Hu, X., Han, Q., 2015. An electrical conductivity model for fractal porous
Burdine, N.T., 1953. Relative permeability calculations from pore size distribution data.
media. Geophys. Res. Lett. 42 (12), 4833–4840.
Trans. AIME 5 (3), 71–78.
Worthington, P.F., Pallatt, N., 1992. Effect of variable saturation exponent upon the
Brooks, R.H., Corey, A.T., 1966. Properties of porous media affecting fluid flow. J. Irrig.
evaluation of hydrocarbon saturation. SPE-20538-PA 7 (4), 331–336.
Drain. Div. 6 (1), 61–88.
Cai, J.C., Wei, W., Hu, X.Y., Wood, D.A., 2017. Electrical conductivity models in saturated

899