ABSTRACT

Many investigation have been carried out to develop either

interfacial tension, IFT, or surface tension relationships.
However, most of them came from exhaustive procedures which require
very specific chemical parameters, namely cationic and anionic
concentrations which are not readily obtained in the Petroleum
Engineering field. Other researchers have looked into the
parameters that the IFT depends on. Many correlations have been
generated. Most of them required such data as molar concentration
and molecular weight, not always available, though. Others have run
some experiments and generated some curves specific for certain
fluids. However, a very general correlation was not found. Besides,
no information relating IFT to well logs was found. It is a need in
the Petroleum Engineering environment to have a "thumb rule" IFT
relationship.
 INTRODUCTION

It has been recognized for many years that surface forces play
an active part in the production history of an oil reservoir as
well as determining the amount of recoverable oil. The magnitudes
of this forces are governed by the values of interfacial tensions,
IFT. These vary greatly with composition, pressure and temperature1.

Interfacial tension is probably the most important single

force that makes one-third of the total oil actually in place in a
reservoir rock unrecoverable
2
by conventional gas-drive or
waterflood methods . It may be defined as the measure of the
specific free
3
surface energy between two phases having different
composition . If the phases have same composition the interfacial
tension is called Surface tension.
Along this study, several papers concerning IFT relationships
have been reviewed. It was found that most of the developed
relationships might result unpractical due to the required data
used by them which are not readily obtainable. Therefore, based on
the literature review, here, it is proposed to make a laboratory
investigation in order to correlate such parameters as pressure,
temperature and density which could be obtain both easier and
quicker.
 3

LITERATURE REVIEW

PAPER I. EFFECT OF PRESSURE AND TEMPERATURE ON OIL-WATER

INTERFACIAL TENSIONS FOR A SERIES OF HYDROCARBONS

M. E. Hassan, R.F. Nielsen and J.C. Calhoun, Oct. 1953

A) PROBLEM BEING SOLVED
To determine the dependency of IFT on pressure and temperature
for benzene, propane, n-pentane, n-hexane, n-octane, and iso-octane
against water, in order to establish a general equation of IFT as a
function of pressure and temperature over their ranges studied.
B) HOW THE PROBLEM WAS SOLVED
The interfacial tension of the named hydrocarbons against
water was measured at temperatures ranging from 26 to 82 °C and at
pressure ranging from 1 to 204 atm. To do so, a device based on an
optical system was constructed.
C) MAIN RESULTS
Five important remarks might be point out from the obtained
data:
i) Several values of IFT were measure and plotted. The deviation
of the data from the fitting curve was about a tenth of a
dyne.
ii) In all cases showed a slight decrease of IFT with pressure at
constant temperature in the range studied. The effect of
pressure becomes less as the pressure was increased with an
indication of a reversal of the effect at higher pressures.
See fig. 1.
iii) There was a decrease of IFT with temperature at constant
pressure in all cases. This rate of decrease became greater
the higher the temperature.
iv) Besides the variation of IFT with pressure and temperature,
there was a higher notable dependency on composition (Compare
fig. 1 and 2).
v) Plotting the results of the isothermal curves against the
logarithm of the pressure indicated a linear relation.
Assuming such a relation, the following empirical equation may
be written:

γ = γ o - aP +bPlog 10P
Where:
γo = is the IFT at atmospheric pressure
a, b = Constants which depend on the system and on the
temperature.
vi) They found a definite trend in the water-oil IFT at given
 4
pressure and temperature with molecular weight for members of
homologous series. Such a trend is indicated by fig. 3. The
data for methane are not strictly compared to those for the
others members of the series (They took the data form methane
and decane from other researchers).

TABLE 1. Constants of equation (1) (t=T-25, Where T=°C)

Compound a b
Propane 0.310+0.0016t 0.100+0.0006t
Butane 0.260+0.0025t 0.085+0.0011t
Pentane 0.110+0.0008t 0.041+0.0002t
Hexane 0.037+0.0014t 0.012+0.0005t
n-octane 0.032+0.0006t 0.010+0.00027t
i-octane 0.065+0.0003t 0.024+0.0001t
Benzene 0.032+0.0005t 0.012+0.0002t
D) CRITICAL ANALYSES
i) An average correlation which includes all the studied compound
was not given.
ii) The correlation was given for one type of hydrocarbon. There
is no correlation for a mixture of gases.
iii) They did not correlate the dependency of IFT on composition.
iv) The last term on equation (1) is not applicable to vacuum
pressures.
 8
PAPER II. CORRELATION OF SURFACE AND INTERFACIAL TENSION OF LIGHT
HYDROCARBONS IN THE CRITICAL REGION

E. W. Hough and G.L. Stegemeier, Dec. 1961

A) PROBLEM BEING SOLVED

To observe the IFT's behavior of the samples studied as the
pressure approaches its critical value, in order to obtain, from
experimental data, surface and interfacial tension of light
hydrocarbons and of two-component hydrocarbons mixtures as a
function of the reduced temperature.
B) HOW THE PROBLEM WAS SOLVED
Extensive experimental work on single component and binary
systems is the basis for the correlations outlined in the studied
paper. These researchers started with the tension-density
difference relationship leading to the Sudgen's parachor:

M γ 3/11
P=
∆ρ

P is the parachor parameter and M is the molecular weight.

From this equation, new values of parachor were calculated for
some of the normal hydrocarbons, using data from previous works.
Fig. 4 shows a regular increase in the parachor with an increase in
molecular size from 1 to 20 carbon atoms.
For two-component systems, they used a modified Weignaug-Katz
correlation (the original one exponent 1/4) which is:
11
⎧⎪ 1 ⎛ X 1 Y1 X 2 - Y 2 ) ⎞ ⎫⎪
3
γ =⎨ ⎜ P1( - ) + P 2( ⎟⎬
⎩⎪ 62.43 ⎝ V x V y V x V y ⎠ ⎭⎪

They compared the values from eq. (2) against not also
experimental data on the methane-normal pentane and methane-normal
decane systems but also other correlations obtaining good approach
(see fig. 5). From their observations, they suggested a generalized
equation for multicomponent systems as follows:
11
⎧⎪ 1 ⎡ ⎛ X 1 Y 1 ⎞ ⎛ ⎞ ⎛ ⎞ ⎤ ⎫⎪ 3
γ =⎨ ⎢ P1 ⎜ - ⎟ + P 2 ⎜ X 2 - Y 2 ⎟ + P n ⎜ X n - Y n ⎟⎥ ⎬
⎪⎩ 62.43 ⎣⎢ ⎝ V x V y ⎠ ⎝Vx Vy⎠ ⎝ V x V y ⎠ ⎦⎥ ⎪⎭

C) MAIN RESULTS
1) New parachor numbers were calculated for normal paraffin
hydrocarbons.
 9
2) Since the properties of the liquid and gas at the critical
point are identical, the line critical is a line of zero IFT
tension; hence, the subsequent iso-IFT lines must roughly
parallel this boundary.
3) Experimental data for two-binary systems were approximated by
the modified Weignaug-Katz correlation.
4) The binary methane-normal pentane system and methane-normal
decane system have been investigated throughout a wide
pressure and temperature range, particularly in the region
near the critical point.
D) CRITICAL ANALYSES
1) There is a proposed correlation for multicomponent systems
whose applicability was not demonstrated.
2) For both eq. (2) and (3) the maximum parachor given was for
normal eicosane.
3) The calculation requires knowledge of the composition of both
phases and the molal volume or density of each phase; which is
not always available.
 12
PAPER III. CORRELATION OF INTERFACIAL TENSION OF HYDROCARBONS

E. W. Hough and H. G. Warren, Dec. 1966

A) PROBLEM BEING SOLVED
To develop a more extended IFT correlation based on
experimental data and previous works.
B) HOW THE PROBLEM WAS SOLVED
Hough and Warren started using eq. (2)

11

⎪⎧ 1 ⎡ ⎛ X 1 Y 1 ⎞ ⎛ X 2 Y 2 ⎞ ⎤ ⎪⎫ 3
γ =⎨ ⎢ 1⎜
P - ⎟ + P 2⎜ - ⎟⎥ ⎬
⎪⎩ 62.43 ⎣ ⎝ V L V V ⎠ ⎝ V L V V ⎠ ⎦ ⎪⎭

For each temperature and pressure at which an experimental value of

IFT was measured, the values of X1, Y1, XL, YV were found and the
following values were calculated:

1 ⎛ X 1 Y1 ⎞
X 1= ⎜ - ⎟
62.43 ⎝ VL V V ⎠

1 ⎛ (1- X 1) (1- Y 1) ⎞
X2= -
62.43 ⎜⎝ V L VV ⎠
⎟

Then, eq. (2) becomes:

γ = {P1 X 1 + P 2 X 2}k

1
Being k=11/3. Letting Y = γ k eq. (4) becomes :

Y = P1 X 1 + P 2 X 2

For each experimental value, the unknown data in eq. (4) were
P1, P2, and k. To solve for the unknown values, various values of k
were selected, experimental values of IFT were obtained and Y was
calculated. P1 and P2 were calculated by applying the method of
least squares to eq. (5). By varying k, the best values of P1 and P2
were found for each system which IFT data were available.
They worked on seven different binary systems, five of them
 13
were taken from information available in the literature (methane
against n-propane, n-butane, n-pentane and n-decane. n-butane
against carbon dioxide). The methane-normal heptane and ethylene-
normal pentane were investigated by Hough and Warren.
C) MAIN RESULTS
1) the parachor and k values obtained by other researchers agreed
quite well with the varies carried out in the investigation.
2) Two new binary systems were investigated.
3) IFT for binary hydrocarbon systems was found to be a direct
function of liquid-vapor phase density difference in the
manner as a single-component system.
D) CRITICAL ANALYSES
1) Although the investigation carried out was very good, no new
correlation was proposed.
2) Information for two systems was just produced.
3) The calculation requires knowledge of the composition of both
phases and the molal volume or density of each phase; which is
not always available.
 14
PAPER IV. CORRELATION OF CAPILLARY NUMBER RELATIONSHIP
FOR SANDSTONE

I. Chatzis and N.R. Morrow, Oct. 1984

A) PROBLEM BEING SOLVED

1
To develop a capillary number relationship for displacement of
oil from water-wet sandstones having permeabilities that varied
over a wide range of values.
B) HOW THE PROBLEM WAS SOLVED
i) MATHEMATICAL MODEL
The ratio of viscous to capillary forces is called the
«capillary number». More a dozen expressions have been used in the
literature to express this ratio.

vµ
N C,1 =
σ

K w∆P
N C,2 =
Lσ

K a∆P
N C,3 =
Lσ

Where v is the darcy velocity of the fluid, µ is its

viscosity, σ is the interfacial tension, ∆P/L is the imposed
gradient across the sample of length L, Ka,w are the specific
permeabilities of the sample to air and water, respectively.
A modified Darcy's equation for linear flow in a core is:

∆P
v = K rw K w
µ L

An expression for capillary number is obtained from eq. (6):

vµ ∆P
= K rw K w
σ Lσ
1
Ratio of viscous to capillary forces.
 15
The pressure drop required to mobilize an oil blob is:

⎛ 1 1⎞
∆ P m = 4σ ⎜ - ⎟
⎝ De D ⎠

The pressure difference required to mobilize the blob is

proportional to the difference in drainage and imbibition capillary
pressures, so:
1 1
P dr - P imb = 4σ ( - )
De D

Considering the contact angle hysteresis;

cos θ R cos θ A
P dr - P imb = 4σ (
- )
D D
Being θR and θA the e
receding and
advancing contact angles, respectively. Dividing both sides by L,
letting λ=L/De and β=D/De, after rearranging, it yields:

∆ P m D e2 4 ⎛ cos θ A ⎞
= ⎜ cos θ R -
Lσ λ⎝ β ⎟⎠

For cubic networks of capillary tubes of diameter De, the

absolute permeability is given by:

φ 2
K = DE
96

Approximating ∆Pm/Lσ with ∆P/Lσ and De² with 96K/φ, the

capillary number equation relating K∆P/Lσ to the reduced oil
saturation is:

∆P 4φ cos θ A
= ( cos θ R - )
Lσ 96 λ β

Thus, the capillary number curve relating K∆P/L to reduced oil

saturation will be unique for geometrically similar systems that
have the same wetting behavior. Variation in shape of the capillary
number will reflect variation in pore geometry indicated by the
values of φ, β and λ.

ii) EXPERIMENTAL PROCEDURE

The porosity was determined in each sample. Several values of
kw and ka were measured at several applied ∆P/L across the sample.
In all experiments, initial and reduced oil saturation
 16
determinations were based on gravimetric material balance. The
samples were highly saturated with oil and then flooded to residual
oil at low capillary number using different pressure drops in order
to have several measurements. The displacing substances were of
known IFT.
C) MAIN RESULTS
1) Typical capillary number curves are illustrated by data
plotted in fig. 6. Note that there is no change in the
residual oil saturation until a critical value of capillary
number is reached.
2) For the sandstone samples investigated, the specific
permeability Kw and Ka were found to be related linearly. The
correlation, which originally is given by a graph was
converted in an equation.

9
Kw= Ka
16

3) (∆P/Lσ)cr is inversely proportional to the permeability as was

predicted for the theoretical considerations that led to
equation (8). Those values of (∆P/Lσ)cr obtained from the
experiments were plotted against ka and ka/φ respectively on
log-log paper from where the final correlations were obtained.
See fig. 7.

∆P
= 32300 K a -0.995
Lσ
−1.15
∆P ⎛ ⎞
= 40900 ⎜ K a ⎟
Lσ ⎝ φ ⎠

D) CRITICAL ANALYSES
1) To find the IFT is necessary to know Kw, ∆P and L.
2) The correlation is not general. There is no consideration when
Kw can vary with the composition of the aqueous phase.
3) The correlation can not be applied to porous media of
distinctly pore structure and possibly very different
weighting conditions.
 19
PAPER V. A NEW MULTICOMPONENT SURFACE TENSION CORRELATION BASED
ON SCALING THEORY

S. Lee and M.C.H. Chien, Apr. 1984

A) PROBLEM BEING SOLVED

To develop an improved surface tension correlation based on
the scaling theory using data and equations from earlier
investigations.
B) HOW THE PROBLEM WAS SOLVED
For pure component
Surface tensions almost invariably decrease with increasing
the temperature, they proposed that;

σ _(1- T r )θ

Where Tr is the reduced temperature, T/Tc, and θ is a constant

equal to 11/9. To become expression (9) in a equation:

σ = Ac(1- T r )θ

Ac is a corresponding-state parameter.

2 / 3 1/ 3 ⎛ 0.1207T br ln P c ⎞
Ac = P c T c ⎜ - 0.16 ⎟
⎝ 1- T br ⎠

Where Tbr is the reduced boiling point.

From the scaling theory, the IFT was expressed in terms of

phase densities which are readily measurable;

( ρ L - ρ V ) − (T c - T )5/16

Becoming expression (11) in an equation:

( ρ L - ρ V )= B( T c - T )
5/16

Before substituting eq. (12) into eq. (10), the physical

nature of B.

Near the critical point the density difference is a function

not only of T-Tc but also of molecular weight, Tc, Pc, and Vc. So,
after applying dimensional analysis, it yields:
 20

(ρ L - ρ V )
= B (Tc - T )5 /16
ρc

B = C Z cn
C is a dimensionless constant, Zc is the compressibility
factor at the critical point and n is an exponent to be fitted.
The Surface tension equation for pure Component and Mixture systems
Substituting eq. (14) and (12) into eq. (10) and rearranging, it
yields:

σ 3.911 = ¿P¿ ( ρ L - ρ V )

P is a new parachor defined as follows;

¿P¿= A1/3.911
c V CoverB
_ _
ρL and ρV are the molal densities.

From a modified way of the Wignaug-Katz correlation, eq. (2)

can be written as:
σ ξm = ρ L [ P L ] - ρ V [ PV ]
Being ξ any exponent. The parachors are obtained by summing
the individual parachors weighted by their molar concentration.
An alternative approach for getting the mixture bulk-phase
parachor is proposed by assuming that the pure component parachor,
eq. (16), can be extended for mixtures, thus:

V cp
¿P p¿= A1/3.911
cp
Bp
The suffix p stands for phase. The parameters Ac, B, and Vc
are obtained by summing the individual ones weighted by their molar
concentration. Therefore, the resulting surface tension equation
for mixture systems is obtained from eq. (15) as follows;

σ 1/3.911
m = ρ L [PsubL] - ρ V [ PV ]
Determination of Parameter B in the Parachor expression
For normal paraffins with carbon number less than and equal to
10, a convenient empirical correlation was developed:

B = 1.8544267 Z -0.52402
c

C) MAIN RESULTS
1) The parachor correlation was tested against other two
correlations giving the less deviation values.
 21
2) The surface tension correlation developed was tested against
45 set of measured binary surface tension data and four sets
of multicomponent oil surface tension data, getting an average
deviation of 3.71 % which is about 50% smaller than the
deviation of the correlation of Weinaug-Katz and Stegmeier.
3) It is worth noting that the surface tension correlation does
not contain adjustable parameters. It is also completely
compatible with existing compositional numerical simulators.
D) CRITICAL ANALYSIS
1) In order to apply the developed correlation is necessary to
know some thermodynamical and physical fluid properties which
are not always available.
 22

RESEARCH PROPOSAL
A) STATEMENT OF THE PROBLEM
Several correlations have been developed to predict either the
interfacial tension or surface tension. However, most of them
utilize many data which are not always available, and therefore,
they might result unpractical. Besides, none of the studied
correlation relates the IFT to well logging.
B) OBJECTIVE OF THE STUDY
Although, any study of this nature must be done based on
experimental data, here it is proposed some ideas in order to
develop a practical IFT relationship based on the aspects
considered in the literature review.
C) PROPOSED APPROACH
IFT depends on pressure, temperature and composition of the
fluids. The last one is the parameter which mainly affects the IFT.

γ = f(P, T, composition)
Those parameters should be related somehow after analyzing
information from experiments.
1. Develop equations
When speaking about composition, one should have in mind
molecular weight, density, volummetric percentages, and salinity
(for the case of formation water). Being the difference of
densities a property readily measurable, a way to relate the
interfacial tension might be:

γ = f(P, T, ∆ρ )
For convenience, it is better to obtain a IFT related to
pressure and density obtained to a specific temperature (i.e. the
standard temperature) and then correct the IFT according to the
appropriate temperature, that is:

γ = f(P, ∆ρ )
γ CORR = γ [f(T)]
A theoretical way for relating parameters in expression (17)
should be done by means of a dimensional analysis, thus:

f( γ , P, ∆ρ )= 0

Using the Π theorem of Buckingham, one have:

M α M β
Π=( 2
)( 2
) (Mover L 3 )= 0
T LT
 23

For M : α + β = - 1
For L : - β = - 3
For T : - 2α - 2 β = 0
Solving for the unknown parameters: α = 2 and β = 3. So:

Π = ( γ 2 P 3∆ρ )
Solving for IFT;
γ = Π P 3∆ρ
Since the IFT varies inversely with temperature;

1
Π P 3ρ α
Therefore; T
γ = F CORR Π P 3∆ρ overT

Being FCORR is the corrector factor. Π and FCORR might be obtained

from experimental data.
The water density is affected by its salinity. So, another
empirical correlation should be taken in account for that, thus:

ρ w = ρ w f(ppm)
 24

2. Generate and analyze data

No data can not be generated until appropriate experiments be
run, in order to fit a true "thumb rule" correlation.
D) DISCUSSION
When generating a correlation relating pressure, temperature,
density difference and salinity (if water is present) -like that
one expressed by equation (18)- is very practical to generate IFT
data using information from either core or well logging. From well
logging, the density, salinity and temperature are easily
measurable.
 25

NOMENCLATURE

a Parameter depending on the system and temperature

b Parameter depending on the system and temperature

D Diameter of interface

De Diameter of capillary restriction

FCORR Correction factor
IFT Interfacial tension

Ka Specific permeability of a sample to air

Kw Specific permeability of a sample to water

L Core length
M Molecular weight
P Pressure or parachor parameter

ppm Salt concentration in water in part per million

T Temperature

v Darcy velocity
V Molal volume

X Molar fraction in the liquid phase

Y Molar fraction in the vapor phase

∆P Pressure difference or drop

∆ρ Density difference

γ IFT

φ Porosity of the core sample

ρ Density
 26
_
ρ Molal density = ρ/M

µ Viscosity

σ Surface tension

θ Contact angle
 27

SUFFIXES

1 Component 1

2 Component 2
A Advanced

a Air

br Reduced boiling point

c Critical or capillary

CORR Correction

L Liquid
m Mixture
n Component n

o Initial or atmospherical conditions

p Phase
R Receding
r reduced
V Vapor

w Water

x Liquid
y vapor
 28

REFERENCES

1. Hassan, M.E., Nielsen, R.F. and Calhoun, J.C. " Effect of

Pressure and Temperature on Oil-Water Interfacial
Tension for a Series of Hydrocarbons ". Petroleum
Transactions, AIME (1953). Vol. 198. pp. 229-302.

2. Hough, E.W. and Steigmeier, G.L. " Correlation of Surface and

Interfacial tension of Light Hydrocarbons in
the Critical Region ". Soc. Pet. Eng. Jour. (Dec., 1961)
pp. 259-263.

3. Hough, E.W. and Warren H.G." Correlation of Interfacial

tension of Hydrocarbons". Soc. Pet. Eng. Jour.
(Dec., 1966) pp. 345-349.
4. Chatziz, I. and Morrow, N.R. " Correlation of Capillary Number
Relationships for Sandstones ". Soc. of Pet.
Eng. of AIME (Oct. 1984). pp. 555-562.

5. Lee, S. and Chien M.C.H. " A new Multicomponent Surface

Tension Correlation based on Scaling Theory ". Paper
SPE/DOE 12643 presented at the SPE/DOE 4th symposium on
EOR held in Tulsa, Ok. April 15-18, 1984. pp. 147-155.