Journal of Petroleum Science and Engineering 124 (2014) 222–231

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

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

Monitoring the influence of dispersed nano-particles on oil–water

relative permeability hysteresis
Mohammad Parvazdavani a,n, Mohsen Masihi b, Mohammad Hossein Ghazanfari b
a
Reservoir Studies Center, Faculty of Up-Stream, Research Institute of Petroleum Industry, Tehran, Iran
b
Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran

art ic l e i nf o a b s t r a c t

Article history: In recent years, polysilicon nanoparticles are used to enhance the oil recovery through the water
Received 22 May 2014 injection process in oilfields. The contributing mechanisms are the reduction of interfacial tension and
Accepted 8 October 2014 wettability alteration which lead to improving or decreasing the oil phase relative permeability and can
Available online 31 October 2014
be traced by change of relative permeability curves. However, profound understanding of the effect of
Keywords: dispersed nano-silica particles on the hysteretic behavior of relative permeability curves remains a
water flooding controversy topic in the literature.
nano-silica particles The current study illustrates the influence of dispersed silica particles on hysteretic trend of two-
relative permeability curves phase curves of oil–water relative permeability. Displacement tests of light crude oil by water as well as
hysteresis
water dispersed Nano-silica particles were performed on a sandstone rock sample, and the relative
permeability curves of oil and water phases were considered for two successive cycles of imbibition and
drainage processes. Primarily, the results revealed that the degree of hysteresis in two phase relative
permeability decreased as the dispersed nano-silica particles were used in the tests; besides the
hysteretic behavior of relative permeability curves decreased as the injection proceeded into its second
cycle. Moreover, the relative permeability of non-wetting phase changed significantly while the changes
for wetting phase did not appear quite considerable. The results of this work can help with finding an
appropriate relative permeability curves during EOR processes in case of changing the direction of
displacement.
& 2014 Elsevier B.V. All rights reserved.

1. Introduction surface leading to changes in the rock wettability level. It can be

Nowadays, nanoscience and technology has been used in oil
and gas industry for improving oil and gas recovery. There are
many areas in which nanotechnology can contribute to raise the
efficiency of production, by making it less expensive and more
environmentally attractive than those that are readily available
(Sunjay, 2011). The nano-fluids are made by addition of nano-
particles to fluids with the purpose of intensification and improve-
ment of some features at low concentrations of the dispersing
medium which leads to the following advantages: increase in
sedimentation stability and creation of thermal, optical, stress–
strain, electrical, rheological and magnetic properties that strongly
depend on the size and shape of the nano-particles (Romanovsky
and Makshina, 2004). SiO2 nano-powder is a new type of aug-
mented injection agent that has the ability of stronger hydro-
phobicity and lipophilicity, and can be absorbed on the rock
n
Corresponding author. Tel.: þ 98 21 66166425; fax: þ 98 21 66022853.
E-mail addresses: parvazdavanim@ripi.ir (M. Parvazdavani),
masihi@sharif.edu (M. Masihi), ghazanfarii@sharif.edu (M.H. Ghazanfari).
classified into two types: lipophobic and hydrophilic polysilicon
nanoparticle (LHPN) and hydrophobic and lipophilic polysilicon
nanoparticle (HLPN) (Ju et al., 2002; Ju and Fan, 2009; Suleimanov
et al., 2011). On the other hand, it can reduce the Inter Facial
Tension (IFT) between two phases, enhance oil effective perme-
ability and reduce injection pressure and augment injection rate
(Wang et al., 2010). One of the effective methods applied to trace
the dominant mechanism during displacement of such fluids is to
extract the relative permeability curves.
There are two basic approaches for extracting relative perme-
ability curves from laboratory core flood tests. In the steady-state
method, the fluids are injected simultaneously into core plug
samples. In the unsteady-state method, a fluid is injected to
displace another fluid already available in the core. Processing
the data of steady-state test is relatively simple, but its experi-
ments are time consuming. On the contrary, unsteady-state
laboratory tests can be performed rapidly, but data processing is
a much more difficult task. Moreover, the steady state methods
have been occasionally used in determination of three-phase
relative permeability curves (Masihi et al., 2011); while, it is more
common to use unsteady state displacement experiments in two

http://dx.doi.org/10.1016/j.petrol.2014.10.005
0920-4105/& 2014 Elsevier B.V. All rights reserved.
 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231 223

Nomenclature PV volume of injection based on pore volume of

core sample
S wc connate water saturation, percent So,w saturation of oil/(nanofluid or water) phase
σ surface tension, N/m2 So,w(min) minimum saturation of oil/(nanofluid or water) phase
de deduction of droplet shape, m So,w(max) maximum saturation of oil/(nanofluid or water) phase
ρBrine density of brine, kg/m3 K ro;w relative permeability of oil/(nanofluid or water) phase
ρCrudeOil density of crude oil, kg/m3 K ro;w ðminÞ minimum relative permeability of oil/(nanofluid or
S2 Saturation in the core outlet, percent water) phase
fo fractional flow of oil, fraction K ro;w ðmaxÞ maximum relative permeability of oil/(nanofluid or
Wi cumulative invading-phase injection volume, m3 water) phase
Ir injectivity capacity, fraction K ro;w ðnormÞ normalized relative permeability of oil/(nanofluid or
u flowing velocity, m/s water) phase
ΔP pressure drop, kPa λ exponent of capillary pressure curve
μw viscosity of water, cP Pd displacement threshold pressure defined from capil-
μo viscosity of oil, cP lary pressure curve

phase systems. There are three approaches in two phase relative
permeability calculations from unsteady state displacement experi-
ments (Hussain et al., 2010; Parvazdavani et al., 2013):
(a) Analytical methods (Johnson et al., 1959; Toth and Civan,
2002); (b) Semi-analytical methods (Civan and Donaldson, 1989;
Udegbunam, 1991); (c) History matching method (Archer and Wong,
1973).
However, at the absence of gravity effects and lack of capillary
dominancy, the Johnson, Bossler, Naumann (JBN) method can be
selected due to its simple and quick calculating procedure.
Since the mechanisms activated by the use of nanoparticles
have their dominant effects whenever injected via carrying fluids
such as water in successive and continuous cycles of injections, the
study of the hysteretic behavior of relative permeability curves are
vital for clarifying two phase transport phenomenon. Curves
related to hysteretic behavior of relative permeability and non-
monotonic change of phase saturations have long been recognized
in almost every process of oil recovery (including water flood in
heterogeneous reservoirs, WAG injection, water coning, gravita-
tional separation, steam and polymer huff-and-puff stimulations).
Numerous experimental and mathematical modeling studies have
focused on the examination of the two-phase history dependent
flow (Braun and Holland, 1995). The hysteresis for the wetting
phase is believed to be very slight and the relative permeability
curve for the wetting phase is approximately similar for both
Firstly, nano-particles were dispersed in the saline water phase
using ultrasonic radiation with 400 W of power for 15 min. The
concentration of the nano-particles was selected 0.1% (i.e. 1000 ppm)
which was reported as the optimum concentration (Safari and
Jamialahmadi, 2013). This content of nano-particles was solved in
brine with 1% concentration (i.e. 10,000 ppm). Properties of nano-
silica particles are given in Table 2.
2.1. Core sample properties
During the experiment, non-fractured (i.e. sand stone) core
sample was used. The properties of the core sample with absolute
permeability of 10.59 mD are given in Table 3.
2.2. Apparatus
In this experimental work, due to fast implementation, the un-
steady state displacement approach was used. A schematic dia-
gram for the connections in the displacement apparatus is illu-
strated in Fig. 1.
Table 1
Working fluids information at standard conditions.

drainage and imbibition processes (Aziz and Settari, 1979). Never- Parameter Value at standard condition
theless, Jones and Roszelle (1987) presented the case of more (101 kPa & 16 1C)
significant difference between imbibition and drainage relative
permeability for water rather than for the oil phase (Jones and API dimensionless 41
Viscosity, μo Pa s 3.25E  3
Roszelle, 1987). None of the previous studies have addressed the
Molecular weight, kg/mol 0.093
effects of nanopowders on hysteretic behavior of two phase flow Density, ρl kg/m3 820
in oilfields by virtue of changing the wettability of reservoir rock C12þ molecular weight, kg/mole 0.250
through their adsorption on porous walls of sandstone samples or C12þ density kg/m3 860
the change of IFT between flowing phases. CO2 viscosity, μg Pa s 0.014E  3
CO2 density, ρv kg/m3 1.869
In the present study, the dominant mechanisms in nano-fluid
injection in the case of EOR processes was examined. By analyzing
the relative permeability curves in both cases of water and nano-
fluid injection, the effect of nano-fluid flooding's on the hysteretic
behavior of relative permeability curves were investigated.
Table 2
Nanosilica properties.

Parameter Value
2. Experimental setup and procedure
Particle dimensionless SiO2
In this study, a light oil sample from one of the oil reservoir Average particle size (m) 14E  9
Specific surface area (BET) (m2/kg) 2007 25E þ 3
from the west Iran was used. Fluid properties are illustrated in Tamped density (kg/m3)  50
Table 1.
224 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231

Table 3 for core sample). To reduce the possibility of capillary end effect in
Core properties. core flood test, the injection rate was changed abruptly and back
flow was done to sweep the remaining water and approach into
Parameters Sandstone
irreducible water saturation at the end of the drainage process.
Length, L m 0.06 After this step, water or nano-fluid was injected at the rate of
Diameter m 0.037 1.67E 9 (m3/s) up to 2 PV and the amount of oil and water or
Area, A m2 0.0011 nano-fluid produced were measured dynamically. Using the pres-
Pore volume, PV m3 7.96E  6
Porosity, φ fraction 0.11
sure drop and produced volumes recorded, the relative perme-
Permeability, K m2 10.59.56E  15 ability curves were extracted. Also, invading-phase saturation was
determined by mass balance calculations. It is emphasized that the
tests done at 5516 kPa injection pressure controlled by BPR and
DP the corresponded overburden pressure was regulated around
10,342 kPa. Moreover, all of the tests had been performed at
ambient temperature, i.e. 23 1C (test temperature).

Core 2.4. Effective mechanisms

HP Air Cylinder
BPR According to the literature, in the low-permeability oilfields,
the augmented injection technology of SiO2 nano-powder can
enhance water capacity for being injected from injection wells and
reduce injection pressure (Wang et al., 2010). However, in the case
Injection Gas
Injection Oil

of permeable core samples just like the one used in this paper, this
mechanism cannot be as effective as expected. To investigate this
issue and find the main mechanisms during such processes, the
Confining Pressure H ydraulic IFT and contact angle tests have been performed and the effects of
such parameters have been investigated.

Interfacial tension test

To measure IFT values, back pressure was set on 5516 kPa as
Pump
well as setting the temperature on test temperature. The pressure
Fig. 1. Schematic diagram of apparatus. drop along the core was about 207–552 kPa. Since the IFT value
was not predicted lower than 5 mN/m, the pendant drop method
was used to measure IFT. The interfacial tension value in each part
Table 4 of water/nanofluid injection was obtained (Fig. 2). The results of
Apparatus specifications. the experiments revealed that the existence of nano-silica without
the use of any additives caused a reduction in the interfacial
Parameters Values
tension.
Working confining pressure kPa Up to 34,354 kPa (option 68,707)
Core sample diameter m 0.0254 (option 0.0381)
Core sample length m 0.02540 to 0.0762 Contact angle test
Type of pump dimensionless HPLC The sessile drop technique is commonly used to measure the
Pressure of pump kPa 0 to 41,225 wetting properties of the smooth, homogenous solid surface
Flow rate of pump m3/s 1.67E 10 to 1.67E 7
regarding two immiscible fluids. It is worthy to mention that the
Confining pressure kPa Up to 68,707
experiments were performed at room temperature (test tempera-
ture). After a static condition is obtained approximately 60 s, the
It consists of an injection system, two piston accumulators, a contact angle was measured, and the observed angle represents a
core-holder, a back pressure regulator, an overburden pressure static contact angle.
system, a pressure differential measurement system, optionally
a video tracker and the computer system for data acquisition and
process control. The injection system consists of a fluid delivery
HPLC pump. The detailed information of apparatus is shown in
Table 4 (Vinci Technologie, 2010).

2.3. Test procedure

The tests consisted of two sets of injection tests, first the

reference water injection and second the nano-fluid injection.
Each of the tests included two successive cycles of injection (i.e.
drainage and imbibition). To start with each run, the core was
cleaned with solvent and dried with hot nitrogen and evacuated.
After core preparation, the core sample was saturated with the
brine of salinity 10,000 ppm by using the saturator apparatus.
After determination of porosity and absolute permeability, oil was
injected into core at the rate 1.67E  9 (m3/s) up to 3 PV (Pore
Volume) in order to reach the irreducible water saturation (0.36 Fig. 2. IFT values before and after nanofluid injection.
 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231 225

Fig. 3. Contact angle values before and after nanofluid injection.

Having equilibrated the two-phase samples, the oil phase and
the aqueous phase were separated, and the cover slips (core slices)
were aged in brine and nano-silica/brine solutions for more than
40 h in 50 1C before measuring the contact angles. Droplets of the
equilibrated oil phase were deposited on cover slips surrounded
by the equilibrated aqueous phase. The oil droplet displaces water
on surface of core slice, giving a dynamic water receding condition
initiated by a water receding angle. As shown in Fig. 3, nano-silica
particles had no dominant effect on change of wettability condi-
tion and IFT change was dominant compared to wettability one.
2.5. Methodology
In order to compute the relative permeability values of each
phase, the JBN method was used. This method is an unsteady state
approach which calculates the relative permeability of the two
phases using the Buckely–Leveret theory, expanded by Welge
(Johnson et al., 1959). There are two main assumptions regarding
the validity of this method. First, the velocity of the fluid flow
should be high enough to reach equilibrium, and second, the
velocity should be constant at any cross section of the core. To
fulfill these assumptions, dimensionless capillary number had
been determined. Based on the calculated dimensionless capillary
number for the injection condition (more than 10  7), the flow in
core flood test is not capillary dominant and the viscous flow
controls the fluid flow behavior (Singh et al., 2001). Additionally,
due to horizontal location of core holder, the no-gravity assump-
tion counts valid. The JBN method calculates relative permeability
curves individually based on Eq. (1). The terms of Wi and Ir are
determined using Eqs. (2) and (3) (Johnson et al., 1959).
dð1=W i I r Þ f ability obtained by history matching, performance of the capillary
¼ o
dð1=W i Þ Kr o
1 1
Wi ¼ 0¼
f df =dS
where fo is considered the fractional flow of oil, Wi is regarded as
the cumulative invading-phase injection volume, Ir is the injectiv-
ity capacity, u is the flowing velocity and ΔP is the pressure drop,
μw is the viscosity of water and μo is the viscosity of oil, us and ΔPs
are the values of u, ΔP is considered at the start of injection. Also
Save and S2 are the average and outlet saturation of invading phase
respectively.
Since the JBN method does not consider the capillary pressure
in calculations, numerical history matching technique was used to
correct the effect of capillary end effect on relative permeability
values specifically at terminal condition (Olafuyi et al., 2008;
Qadeer et al, 1988).
Inverse modeling (i.e. the history matching technique) has been
used by MATLAB software to extract the relative permeability
curves implicitly. The production results obtained with solving the
flow equations of the finite difference model (one dimensional,
horizontal two phase flow with capillary effects) by the IMPES
method.
By comparing the modeling production results with the experi-
mental ones, the initial guess of the relative permeability which
was achieved by using the JBN method, would be modified by
using a suitable optimization algorithm (the conjugate Margaret
method).
Spline parameters of relative permeability curves were con-
sidered as the controlling parameter and absolute difference of
production results (qo as the cumulative oil production or DP as
the pressure drop) obtained by both experimental and simulation
acted as an objective function. By minimizing the objective
function, the ultimate values of relative permeability were con-
sidered as the implicit true values (Parvazdavani et al., 2013).
To evaluate the validity of implicit values of relative perme-
ð1Þ
end effect correction method which corrects the terminal satura-
tion and relative permeability of flow phases by Eqs. (7)–(11)
(Huang and Honarpour, 1996) was compared with the history
ð2Þ
match one.

u=ΔP    ð1=λÞ 
Ir ¼ ð3Þ kroðSiwÞ KkroðsiwÞ AP d
us =ΔP s ¼X Snw; inlet 1 ; X¼ ð7Þ
kro ðave; at Siw ðaveÞÞ Lqo μo

1  f o mw
Kr w ¼ Kr o ð4Þ
f o mo SiwðaveÞ  SniwðaveÞ ð1  Sor Þ
Siw ¼ ð8Þ
1  SniwðaveÞ
Relative permeability curves were determined based on outlet
invading phase saturation which can be counted by Eqs. (5) and
(6) (Johnson et al., 1959). where Sniw ðaveÞ would be calculated by

 2 3
Save ¼ S2 þW i f o 2 ð5Þ 2

2 λ λ þ 2 ð18 λ2
þ λ þ 1Þ
Siw ðaveÞ ¼ 4
n
  
 5X

 ðλ  1Þ 4λ  1 9λ  1 4λ þ 1
2 2
S2 ¼ Save W i f o 2 ð6Þ
226 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231

" n n n
#
2 3 9
n
 XSw; inlet ð  1=λÞ Sw; inlet  2Sw; inlet þ Sw; inlet
λ1 2λ  1 3λ  1 8 Exp.

" n n n
# Sim.
2 3 4
ð  1=λÞ Sw; inlet  2Sw; inlet þ Sw; inlet
7

Volume (Cubic meter)*E-6

þ XSnw; inlet ð9Þ
2λ þ 1 3λ þ1 4λ þ 1 6

where X would be calculated by 5

KkroðsiwÞ AP d 4
X¼ ð10Þ
Lqo μo 3
n
Sw; inlet can be calculated by solving equation numerically. 2
2
 2 
KkroðsiwÞ AP d n 2λ λ þ 2 6λ þ λ þ 1 1
 1=λ
1¼ ½ðSw; inlet Þ  2  

Lqo μo λ  1 4λ 2  1 3λ þ 1 0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2  n 1  ð1=λÞ 1  n 2  ð1=λÞ P.V. (Oil Injected)
þ Sw; inlet  Sw; inlet
λ 1 2λ  1 Fig. 5. History matched cumulative water production curve in 1st drainage process
1 1  n 2 þ ð1=λÞ 1  n 3 þ ð1=λÞ
þ ðSnw; inlet Þ1 þ ð1=λÞ  Sw; inlet þ Sw; inlet of water injection.
λþ1 2λ þ 1 3λ þ 1
ð11Þ
values for successive nano-fluid injection both cycles are shown in
Fig. 7. It should be noted that these curves are free from capillary
end effect especially at terminal conditions (i.e. end point relative
3. Results and discussion
permeability). Since the dominant source of error due to capillary
end effect is in the drainage process (Huang and Honarpour, 1996),
To investigate the effects of history changes on the relative
three main curves are shown in these figures; firstly JBN drainage
permeability curves, two successive injection cycles were con-
relative permeability (including the capillary end effect), secondly
ducted in each process (i.e. water/nanofluid injection). Also by
history matching drainage relative permeability (removing capillary
comparing the relative permeability curves of water flooding
end effect) and imbibition curves. As observed, the most hysteric
displacement with nano-fluid flooding displacement tests, the
behavior in rel perm curves is during the water injection process.
effect of nano-particles additives on the change of relative perme-
Non-monotonic change of phase saturations is due to the difference
ability curves and its hysteretic behavior had been investigated.
between contact angles in case of water and oil advancing respec-
Moreover, in each process, the relative permeability curves of
tively. On the other hand, The main origins of hysteresis is
successive injection cycles were compared to find the dominant
irreversibility of pore filling sequence (type 1 hysteresis) and
phase for which relative permeability curves change more con-
trapping/coalescence hysteresis (type 2 hysteresis). In the first type,
siderably. By considering the raw data (i.e. production and pres-
during imbibition, water invades the smallest throats (except the
sure drop) after the flooding phase breakthrough time, the initial
throats which contain disconnected oil). During drainage, oil tends
guess of relative permeability values were extracted by using the
to invade the largest throats. However, not all large throats are
JBN method to be used in history matching technique. The history
accessible. The irreversibility leads to hysteresis in the so-called
matched production results of the 1st cycle of water injection are
backbone saturation. This effect causes increase in oil relative
shown in Figs. 4 and 5. As shown, the pressure drop curves
permeability curve in case of drainage compared to imbibition. In
decreased after break through time to approach into
the second type, during imbibition, part of the oil is disconnected by
stabilized value.
water and does therefore not contribute to the oil conductivity.
The relative permeability values for both successive water
Most of the trapping occurs close to the oil flow threshold below
injection cycles (i.e. drainage and imbibition curves) which are
which no continuous oil paths exist. In addition, trapping can also
obtained by inverse modeling are plotted in Fig. 6. The same
be a result of snap-off. During drainage, connected oil coalesces
with oil ganglia at a certain rate. This rate is assumed to be
determined by the corresponding throat filling sequence. The oil
75 ganglia are therefore not necessarily released at the same rate as
Sim.
Exp. they are created. Hence, as observed, the capillary end effect has the
most effect at terminal nonwetting relative permeability and
70
irreducible wetting saturation. This effect is due to increase in
water saturation at outlet of core sample during the process of
drainage. History matching technique removes this effect by use of
65
optimization process on relative permeability parameters. In this
DP (Kpa)

process, the capillary pressure was embedded in fluid flow calcula-

60
tions (see capillary pressure for oil and water in Fig. 8) to reduce the
effect of capillary dominancy at the end of the drainage process.
Besides the effect of nanoparticles on hysteretic behavior of fluid
55 flow, the change of relative permeability of flow phases due to
formation damage based on the variation of particle sizes was
investigated. Larger size particles cause pore blocking, medium sizes
50 (slightly less than a pore throat) can cause pore bridging and small
0 0.2 0.4 0.6 0.8 1 1.2 1.4 size (far less than the pore throat) can lead to adhesion of particles to
P.V. (After breakthrough) pore walls. Also, mobilization of fines with permeability decline in
Fig. 4. History matched pressure drop curve in 1st drainage process of water the case of low water salinity has been thoroughly reported in the
injection. literatures. This effect is usually attributed to the expansion of double
 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231 227

1 1
Drainage - end effect Drainage - end effect
0.9 Drainage - end effect 0.9 Drainage - end effect

Drainage - No end effect Drainage - No end effect

0.8 0.8 Drainage - No end effect
Drainage - No end effect
Imbibition
Imbibition
0.7 0.7
Relative Permeability

Relative Permeability
Imbibition
Imbibition
0.6 0.6

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Sw Sw

Fig. 6. Relative permeability curves of the 1st and 2nd cycle of water injection.

1 1
Drainage - end effect Drainge - end effect
0.9 Drainage - end effect
0.9 Drainge - end effect
Drainage - No end effect Drainge - No end effect
0.8 0.8
Drainage - No end effect Drainge - No end effect
Imbibition Imbibition
0.7 0.7
Relative Permeability

Relative permeability

Imbibition Imbibition

0.6 0.6

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Sw Sw

Fig. 7. Relative permeability curves of the 1st and 2nd cycle of nanofluid injection.

70 100
90
60
80
Pore Size Frequency, %

50 70
60
PC (Psi)

40
50
30 40
30
20
20
10 10
0
0 0.001 0.01 0.1 1 10 100
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Sw Pore Radius (micro meter)

Fig. 8. Capillary pressure curve for oil/water drainage process. Fig. 9. Pore size distribution curve for core sample.

electric layers between the fine particle and rock surface, weakening on rock surface (Fig. 10). By qualitative comparison of the figures
of the total grain-rock electrostatic attraction which consequently (after and before nano-silica injection), there is no significant scale
leads to particle detachment by dragging and lifting forces. The regarding nano-particle adsorption on the rock sample surface. To
detachment of fine particles leads to a slight increase in permeability have educational challenge on effect of nano-particle transport on
while the plugging of the small pores causes significant decline in possibility of damage, we should have pore size distribution to
permeability (Bedrikovetsky et al., 2012). compare the average size of particles and throat size. This plot was
To investigate the effect of nano-particles on pore plugging, shown in Fig. 9. As shown, approximately most of the pore throat
Scanning Electron Microscope (SEM) test was performed after and sizes (nearly 90%) are larger than 0.3 mm compared to average size
before the nanofluid injection to track the promising adsorption of nanoparticles (0.014 mm). This shows the higher radius of throat
228 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231
After Injection
Fig. 10. SEM test after and before the nano-silica particle injection.
Fig. 11. Comparison of pressure drop curves of water and nano-fluid injection at
the 1st cycle.
sizes compared to particle size and consequently, lack of possibi-
lity for damage. This fact was approved qualitively with SEM
results (Fig. 10). This behavior can be traced with no-increase in
pressure drop curve during nanofluid injection compared with
brine injection case (Fig. 11). As observed, pressure drop at break-
through in case of nanofluid is higher than the corresponded value
of water injection. This is due to higher oil recovery at inflection
point of oil production curve in case of nanofluid compared to
water injection. At the last part of curve, since there was not any
considerable damage, the pressure drop value is the same for both
nanofluid and water.
To find the more dominant phase in hysteretic behavior of
relative permeability curves, Figs. 6 and 7 were used for successive
drainage and imbibition tests. They revealed that during the non-
wetting-phase (i.e. oil) relative permeability curves have changed
more considerably than wetting phase in hysteretic behavior. To
comprehend the main reason of this phenomenon, we should
understand the fluid flow behavior through porous medium
during drainage and imbibition. In drainage mechanism, the
wetting-phase in large pores throats with high curvature radius
will have been depleted before small pores. This event is reversed
for non-wetted phase in imbibition mechanism. So that the
saturation path alters more in non-wetting phase rather than the
wetting phase in successive cycles of drainage and imbibition that
verified the dominant effect of hysteretic behavior on non-wetted
phase. Also, it can be observed that the hysteretic behavior of
relative permeability curves decreased as the injection proceeded
Before Injection
into the 2nd cycle. This took place because of the establishment of
saturation paths as the injection cycle was carried on.
From Fig. 7A, it can be inferred that the intersection which
happened in non-wetting phase relative permeability curves is
due to lack of similarity to the wetting phase. To be more precise,
the first drainage of nano-flooding was performed in the presence
of brine water and oil, while in the first imbibition process after
drainage, nano-fluid substituted water as the wetting phase.
Hence, in order to understand the effect of nano-fluid on the
hysteretic behavior of relative permeability curves, we should
compare Figs. 6B and 7B. Hysteretic behavior is more dominant
in water flooding (see Fig. 6) in comparison with that of nano-fluid
injection (see Fig. 7). When the nano-particles were added to
water, IFT values reduced (Fig. 2) and this reduction caused low
resistance between the wetting (nanofluid) and non-wetting (oil)
phases and this effect led to the sameness in relative permeability
values of oil/nanofluid phases in drainage and imbibition pro-
cesses and consequently weakened the hysteretic behavior. On the
contrary, as surface absorption is the intrinsic behavior of nano-
particles, the alteration in the wettability of the core sample from
water wet to oil wet is checked and is found quite insignificant
(Fig. 3). As a result, it could be deduced that the mechanism of IFT
reduction is more dominant. The main indicator of this claim is
related to the change of relative permeability curves in the case of
nano-flooding compared with water flooding (Fig. 12). As
observed, in case that the decrease in the IFT is more dominant
in comparison with wettability change, the relative permeabilities
of both phases increase. While whenever the wettability change is
imminent, the relative permeability of non-wetting phase
decreases and wetting phase values increase. Another indicator
is related to the change of pressure drop curve, in case wettability
changes, the rock approaches into oil wet and water break through
occurs sooner than the water wet case. Obtained results show that
the break through time is the same for both cases of water and
nanofluid injection and consequently wettability change is not
dominant (Fig. 10). On the other hand, if surface adsorption of
nano-particles occurs dominantly and wettability changes from
strong water-wet to mixed-wet or oil-wet, the pressure drop along
the core length should increase sharply in case of sweeping oil
phase by the water dispersed nanosilica particles, while this
phenomenon has not been observed through the experimental
results.
Prior to investigate the effect of nanofluid injection on imbibi-
tion relative permeability values as an EOR method, an analysis
was done on removing the capillary end effect from the relative
 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231 229

1
1

0.8
0.8

Relative Permeability
Relative Permeability

0.6
0.6

0.4 0.4

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Sw Sw

Fig. 12. Comparison of relative permeability of water and nano-fluid injection at the 1st and 2nd cycle.

1.8 Table 5
End effect analysis on end point relative permeability values.
1.7
Process End point End point (Kro_corrected/
1.6
(Kro_corrected/ Kro_measured), end
Kro_measured), history effect correction method
Kro(Siw)/Kro-ave(Siw-ave)

1.5
matched method
1.4
1st cycle of water 1.4 1.38
1.3 2nd cycle of water 1.3 1.3
1st cycle of nano 1.34 1.35
1.2 2nd cycle of nano 1.26 1.25

1.1

1
Table 6
0.9
End effect analysis on end point water saturation values.
0.8
10-2 10-1 100 Process End point End point End point
saturation, saturation, saturation, end
Pd/DP
measured history effect correction
Fig. 13. Determination of corrected terminal oil relative permeability at drainage matched method
process.
1st cycle of water 0.37 0.34 0.35
2nd cycle of water 0.43 0.41 0.41
1st cycle of nano 0.37 0.33 0.34
100
2nd cycle of nano 0.49 0.47 0.48

water saturation was plotted versus the described ratio of pressure

10-1 values (Fig. 14). To analyze the validity of history matched relative
permeability values of non-wetting phase bearing the most effect
of capillary dominancy, the ratio of terminal oil relative perme-
ability in case of end effect free (history match or numerical
analysis (see appendix) method) to JBN derived value was shown
10-2
in Table 5. The corresponded values for end point water saturation
for different methods were shown in Table 6.
To compare the results of nano-fluid with water flooding,
Fig. 15 was plotted based on the normalized relative permeability
10-3 versus normalized saturation defined by
10-2 10-1 100
So;w  So;wðminÞ
Fig. 14. Determination of corrected terminal water saturation at drainage process. So;wðnormÞ ¼ ð12Þ
So;wðmaxÞ  So;wðminÞ

permeability values of the drainage process. In this analysis, the K ro;w

ratio of relative permeability at terminal (end effect free) to K ro;wðnormÞ ¼ ð13Þ
K ro;wðmaxÞ
average condition (JBN calculated) was plotted versus ratio of
capillary threshold displacement pressure (Pd) to viscous pressure where So,w(min), So,w(max) and Ko,w(max) refer to the minimum and
drop along the core sample (ΔPl) (Fig. 13). Also the normalized maximum outlet invading fluid (i.e. water or oil) saturation and
230 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231

1 1

0.8 0.8

Relative Permeability (Normalized)

Relative Permeability (Normalized)

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Sw (Normalized) Sw (Normalized)

Fig. 15. Comparison of relative permeability curves in both cases of water and nanofluid injection at the 1st and 2nd cycles of injection in normalized scale.

Table 7 Table 10
Production results of water injection. Uncertainty analysis in successive cycles of water/nanofluid injection.

Process Recovery factor End point saturation Process Krw (error) Kro (error) Error_ave

1st Drainage 0.616 0.37 Comparison of 1st cycle of water/nano 1.89 17.40 17.50
1st Imbibition 0.566 0.27 Comparison of 2nd cycle of water/nano 1.39 13.14 13.21
2nd Drainage 0.421 0.43
2nd Imbibition 0.424 0.25

In order to find the dominant phase in change of relative

permeability regarding hysteresis effect in successive imbibition
Table 8 and drainage processes as well as monitoring the quantitative
Production results of nanofluid injection. effect of nanofluid on hysteresis behavior, the uncertainty analysis
of the obtained results was performed and the corresponding
Process Recovery factor End point saturation
results were shown in Tables 9 and 10. As it was shown in Table 9,
1st Drainage 0.60 0.37
the non-wetting phase was the dominant phase in hysteretic
1st Imbibition 0.72 0.14 behavior and this hysteresis, which is defined with error distin-
2nd Drainage 0.44 0.48 guished by Eqs. (14) and (15) for first and second cycles of
2nd Imbibition 0.71 0.15 injection in each saturation point, diminishes rapidly from first
to second cycle of water injection. Eq. (15) was used to calculate
the error of relative permeability values for injection cycles in each
Table 9 case of nanofluid and water. The final error of relative permeability
Uncertainty analysis in two successive cycles of water and Nanofluid injection. was computed by Eq. (16). The slope of this trend reduced in case
of nanofluid injection and both of first and second cycles of
Process Krw (error) Kro (error) Error_ave
nanofluid injection had the lower value of hysteresis (Error_ave)
1st cycle of water 2.33 25.47 25.58
compared to the second cycle of water injection. Table 8 illustrated
2nd cycle of water 0.85 7.00 7.06 that in the second cycle of nanofluid injection, the hysteresis
1st cycle of nano 0.70 3.83 3.90 between the imbibition relative permeability curves of water
2nd cycle of nano 0.71 2.35 2.46 and nanofluid injection was lower than the first cycle which
implied that the major share of higher oil recovery caused by
the use of nanofluid injection was the result of the first cycle of
maximum relative permeability values of oil/water phases injection.
respectively.  
Kr o;w ðImbibitionÞ  Kr o;w ðDrainageÞ
Additionally, by reviewing the results (i.e. Fig. 12) it can be Kr o;w;Sw ðHysteritic_ErrorÞ ¼
figured that non-wetting phase relative permeability curves chan- Kr o;w ðDrainageÞ
ged more considerably than wetting phase curves in nanofluid ð14Þ
flooding in contrast to water flooding case, which led to enhanced
 
oil recovery in case of nano-flooding. Tables 7 and 8 show the Kr o;w ðNano FluidImbibition Þ  Kr o;w ðWater Imbibition Þ
Kr o;w;Sw ðEOR_ErrorÞ ¼
values of recovery factor in each drainage and imbibition process Kr o;w ðWater Imbibition Þ
of two water and nano-fluid injection cycles respectively. Also, the ð15Þ
end point saturation values of drainage and imbibition processes
(i.e. connate water and irreducible oil saturation) are illustrated. qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
These values are calculated based on the oil and water production Kr o;w ðErrorÞ ¼ ∑ðKr o;w;Sw ðiÞÞ2 for i ¼ EOR or Hysteric ð16Þ
considering dead volumes and pore volumes of core sample.
 M. Parvazdavani et al. / Journal of Petroleum Science and Engineering 124 (2014) 222–231
4. Conclusions
Experimental unsteady state displacement tests of water/water
saturated by dispersed nanoparticles-light crude oil systems
were performed on a sandstone core sample resulting in the
determination of relative permeability curves of both water and
oil phases for two successive cycles of both imbibition and
drainage processes. Based on the results obtained in this experi-
ment, the following conclusions can be reached:
(1) Nano-silica particles have more impact on the change of the
IFT value than the wettability condition. (2) Non-wetting-phase
has stronger changes in hysteretic behavior of relative permeabil-
ity curves in successive processes of drainage and imbibition than
wetting phase in case of nanofluid injection along with brine
injection. (3) The hysteretic behavior of relative permeability
curves decreased as the injection proceeded into the 2nd cycle
of injection. (4) By nanofluid core flooding, the hysteresis in the
relative permeability curves reduced and seemed to be less
compared to that of reference water flooding. (5) Nano-particles
additives have more effect on the change of the non-wetting phase
relative permeability curves than the wetting-phase.
References
Archer, J.S., Wong, S.W., 1973. Use of a reservoir simulator to interpret laboratory
water flood data. SPE J. 13 (6), 343–347.
Aziz, K., Settari, A., 1979. Petroleum Reservoir Simulation. Applied Science Publish-
ers, London WY.
Bedrikovetsky, Pavel, Vaz, Alexandre, Machado, Fernando, Zeinijahromi, Abbas,
Borazjani, Sara, 2012. Skin due to fines mobilisation, migration and straining
during steady state oil production. J. Pet. Sci. Technol. 30, 1539–1547.
Braun, E.M., Holland, R.F., 1995. Relative permeability hysteresis, laboratory
measurements and a conceptual model. SPE Reserv. Eng. 10 (3), 222–228.
Civan, F., Donaldson, E.C., 1989. Relative permeability from unsteady state displace-
ments with capillary pressure included. SPE Form. Eval. 4 (2), 189–193.
Huang, D., Honarpour M., 1996. Capillary end effects in coreflood calculations. In:
Proceedings of SCA Conference. Paper Number 9634.
Hussain, F., Cinar, Y., Bedrikovetsky, P., 24–28 April 2010. Comparison of methods
for drainage relative permeability estimation from displacement tests. In: Paper
SPE: SPE Improved Oil Recovery Symposium. Tulsa, Oklahoma, USA.
231
Johnson, E.F., Bossler, D.P., Naumann, V.O., 1959. Calculation of relative permeability
from displacement experiments. Trans. AIME 216, 370–372.
Jones, S.C., Roszelle, W.O., 1987. Graphical techniques for determining relative
permeability from displacement experiments. J. Pet. Technol. 30 (5), 807–817.
Ju, B., Dai, S., Luan, Z., Zhu, T., Su, X., Qiu, X., 8–10 October, 2002. A study of
wettability and permeability change caused by adsorption of nanometer
structured polysilicon on the surface of porous media. In: Paper SPE 77938,
SPE Asia Pacific Oil and Gas Conference and Exhibition. Melbourne, Australia.
Ju, B., Fan, T., 2009. Experimental study and mathematical model of nanoparticle
transport in porous media. Powder Technol. 192 (2), 195–202.
Masihi, M., Javanbakht, L., Bahaloo Horeh, F., Rasaei, M.R., 2011. Experimental
investigation and evaluation of three-phase relative permeability models. J. Pet.
Sci. Eng. 79 (1–2), 45–53.
Olafuyi O., Cinar Y., Knackstedt M., Pinczewski w., 2008. Capillary Pressure and
Relative Permeability of Small Cores, SPE-113386-MS. Tulsa, USA.
Parvazdavani, M., Masihi, M., Ghazanfari, M.H., 2013. Gas–oil relative permeability
at near miscible condition: an experimental and modeling approach. Sci. Iran
20 (3), 626–636.
Qadeer, S., Dehghani, K., Ogbe, D.O., Ostermann, R.D., 1988. Correcting Oil/Water
Relative Permeability Data for Capillary End Effect in Displacement Experi-
ments, SPE-17423-MS. California, USA.
Romanovsky, B.V., Makshina, E.V., 2004. Nano-composites as functional materials.
Soros Educ. J. Russ. 8 (2), 50–55.
Safari, M., Jamialahmadi, M., 2013. Effect of polysilicon nanoparticles on enhanced
oil recovery in Iranian oil reservoir (Winter). Int. J. Nano Dimens. 3 (3), 199–205
(SSN: 2008-8868).
Singh, M., Mani, V., Honarpour, M.M., Mohanty, K.K., 2001. Comparison of viscous
and gravity dominated gas–oil relative permeabilities. J. Pet. Sci. Eng. 30 (2),
67–81.
Suleimanov, B.A., Ismailov, F.S., Veliyev, E.F., 2011. Nanofluid for enhanced oil
recovery. J. Pet. Sci. Eng. 78 (2), 431–437.
Sunjay, J., 2011. Nano-Science & Technology in Upstream, Biennial International
Conference and Exposition on Petroleum Geophysics. Hyderabad.
Toth, T. Bodi, Civan, F., 2002. Convenient formulae for determination of relative
permeability from unsteady-state fluid displacements in core plugs. J. Pet. Sci.
Eng. 36 (1–2), 33–44.
Udegbunam, E.O., 1991. A FORTRAN program for interpretation of relative perme-
ability from unsteady-state displacements with capillary pressure included.
Comput. Geosci. 17 (10), 1351–1357.
Vinci Technologie, 2010. Operating Manual of BRP Apparatus. VINCI Technologies,
France.
Wang, K.L., Cheng Liang, S., Wang, C.C., 2010. Research of improving water injection
effect by using active SiO2 Nano-Powder in the low-permability oil field.
Advanced Materials Research 92, 207–212.