SPE 100059

A Single-Shot Method for Capillary Pressure Curve Measurement Using Centrifuge

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

and Quantitative Magnetic Resonance Imaging
Q. Chen, SPE, and B.J. Balcom, U. of New Brunswick

Copyright 2006, Society of Petroleum Engineers

approximately 10 times less than the traditional method. Since
This paper was prepared for presentation at the 2006 SPE/DOE Symposium on Improved Oil only a single moderate rotational speed is employed, the
Recovery held in Tulsa, Oklahoma, U.S.A., 22–26 April 2006.
outflow boundary condition can be maintained, and the effect
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
of gravity can be neglected. In addition, the long rock cores
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to employed for the single-shot method result in a relatively
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at small radial effect.
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is INTRODUCTION
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous Capillary pressure results from the interaction between a
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
wetting fluid, and a non-wetting fluid, as well as their
bounding solid matrix. Capillary pressure critically influences
Abstract the initial reservoir fluid distribution and dynamic processes of
The centrifuge method for capillary pressure curve oil recovery. Capillary pressure is the most fundamental rock-
measurement involves increasing the centrifuge speed in steps fluid property in multi-phase flows, just as porosity and
and measuring the liquid expelled from a short core plug, at permeability are the most fundamental properties in single-
equilibrium, for each step. However, the traditional methods phase flow in oil and gas reservoirs [1].
for deducing approximate solutions for the capillary pressure In evaluating hydrocarbon reservoirs, laboratory
curve are based on the assumption that the capillary pressure capillary pressure curve measurements on reservoir cores are
is zero at the outflow end of the core. In addition, the directly applied to determine many basic petrophysical
traditional centrifuge methods for capillary pressure properties, for example: pore size distribution, irreducible
measurement are time consuming. A full capillary pressure water saturation, residual oil saturation, and wettability of
curve requires approximately 10 different rotational speeds. reservoir rocks. In addition, they are used to determine the
We have observed for most sedimentary rocks that the initial water and oil saturation as a function of height above
experimental magnetic resonance free induction decay is the free water level, approximate oil recovery efficiency, and
single exponential and the effective transverse relaxation time to calculate the relative permeability [2-4]. Capillary pressure
(T2*) is largely insensitive to fluid saturation. These features can also have a significant impact on water flood performance
ensure that Centric Scan SPRITE (single-point ramped [5].
imaging with T1 enhancement) is a quantitative magnetic In the laboratory, the capillary pressure curve can be
resonance imaging (MRI) method, since its local image determined with (1) mercury injection method, (2) porous
intensity is directly proportional to the local fluid content. We diaphragm or membrane, and (3) centrifugal methods, based
propose a single-shot method to measure the capillary pressure on hydrostatic equilibrium [6]. The porous diaphragm method
curve of a long rock core using a single-speed centrifuge is a direct and accurate technique, but the measurement is
experiment and one-dimensional Centric Scan SPRITE MRI extremely time-consuming, since the equilibration time can be
to determine the fluid saturation distribution, S(r), along the weeks or months for each individual pressure point. The
length of the core. A full capillary pressure curve can be mercury injection method is rapid, but it is a destructive
directly determined by the relation of S(r) and the capillary method. In addition, the mercury injection measurement does
pressure distribution, Pc(r), along the length of the core. not provide information on reservoir wettability, and mercury
The single-shot method, employing a desktop is hazardous to human health and the environment.
centrifuge and a desktop permanent magnet based one- The centrifuge method for capillary pressure curve
dimensional MRI instrument, has been applied to measure the measurement was first introduced by Hassler and Brunner [7]
primary drainage, imbibition, and secondary drainage in 1945. This method involves rotating fluid bearing rock
capillary pressure curves for reservoir rocks. cores at variable rotational speeds in a specially modified
The proposed method for determining the capillary centrifuge. The rotation of the sample yields a centrifugal
pressure curve is rapid, cheap, and precise. The capillary force which will counterbalance the capillary pressure when
pressure curve can be obtained straightforwardly with about hydrostatic equilibrium is reached. Collecting and measuring
40 data points. The duration of the experiment is the expelled fluid after hydrostatic equilibrium as a function of
2 SPE 100059

the rotational speeds permit a quantification of the capillary length of the core was negligible compared to the radius of
pressure as a function of fluid saturation [8-10]. A full rotation.
capillary pressure curve determined with traditional
In a centrifuge capillary pressure experiment, a fluid
centrifugal methods requires approximately 10 different
saturated core plug, confined in a special core-holder, is
centrifuge speeds, thus the measurement for a capillary
rotated at different rotational speeds. The relevant distances
pressure curve takes several days to more than a week
are denoted as r1, r, and r2, as illustrated in Figure 1, where r1,
depending on equilibrium time. In addition, the experiment
r2, and r are the distances from the rotational axis to the inlet
requires a very expensive ultracentrifuge with precise speed face, the outlet face, and any point along the core length,
control over a wide range of speeds. This technique has been

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

respectively. The core-holder contains another fluid which
extensively investigated, and is commonly used in the
replaces the fluid displaced from the core. After reaching
petroleum industry.
hydrostatic equilibrium, the amount of liquid expelled from
In the current investigation, we propose a “single-
the core plug is measured with a stroboscope [10].
shot” method to measure the capillary pressure curve using a
single-speed centrifuge experiment and one dimensional
quantitative magnetic resonance imaging (MRI) to determine
the fluid saturation distribution, S(r), along the length of the
sample, instead of non-quantitative spin echo based MRI [11],
and nuclear tracer imaging techniques [12]. A full capillary
pressure curve can be determined by the relationship between
S(r) and the capillary pressure distribution, Pc(r), along the
length of the core [13]. Figure 1. Schematic of a rock core plug spinning in a centrifuge.
MRI is a non-destructive measurement method,
which, with quantitative imaging techniques such as SPRITE The basic concepts for capillary pressure
(single-point ramped imaging with T1 enhancement) [14], measurement with a centrifuge are outlined below.
offers unique advantages in determining spatially resolved When a cylindrical core is placed in a centrifuge, a
fluid saturation in porous rocks [15, 16]. centrifugal acceleration ac = -ω2r, is generated, where ω is the
angular rotation speed of the centrifuge and r is the distance
CAPILLARY PRESSURE THEORY from the axis of rotation. Applying Darcy’s law at hydrostatic
equilibrium, we have
Capillary Pressure dPc
Two immiscible fluids (non-wetting phase and wetting phase) + Δρω 2 r = 0 (4)
in contact in the pore network of a porous medium, produce a
dr
discontinuity in pressure across the interface [17]. The where Δρ is the density difference between wetting fluid and
pressure difference across the interface is called the capillary non-wetting fluid.
pressure, Pc, which is defined by the following equation The differential equation can be solved by simple
integration
Pc = Pnon−wetting − Pwetting (1)
The capillary pressure is a result of the curvature of 1
the fluid interfaces, and the interfacial (or surface) tension, Pc (r ) = Δρω 2 (r22 − r 2 ) + Pc 2 (5)
according to the Young–Laplace equation [5]: 2
⎛1 1 ⎞ If we apply the Hassler-Brunner boundary condition,
Pc = σ ⎜⎜ + ⎟⎟ (2)
i.e., the capillary pressure at the outlet face of the core is
⎝ R1 R2 ⎠ assumed to be zero, i.e., Pc2 = 0, so
where σis the interfacial (or surface) tension between the two
fluid phases and R1 and R2 are principal radii of curvature.
1
If the pore throat shape of a porous medium can be Pc (r ) = Δρω 2 (r22 − r 2 ) (6)
described as a cylindrical capillary tube, equation (2) becomes 2
Pc = 2σ cosθ / R (3)
and for a continuous phase, the capillary pressure at the inner
where R is the pore throat radius, and θ is the contact angle. face of the core is
The capillary pressure can be converted directly into
a pore throat size according to equation (3).
1
PcL (r ) = Pc (r1 ) = Δρω 2 (r22 − r12 ) (7)
Traditional Centrifuge Methods 2
In 1945, Hassler and Brunner [7] proposed a centrifuge The capillary pressure distribution results in a fluid
method to determine capillary pressure-saturation data from saturation distribution along the length of the core. Neither of
small core plugs. They also proposed an approximate solution these distributions is actually measured with the traditional
to the basic equation relating capillary pressure and average method. What is measured is the rotational speed, ω, and the
saturation by neglecting the gravity effect and assuming the average fluid saturation, S , within the core.
SPE 100059 3

The average fluid saturation of the core after from '0' to '-kz', corresponding to a gradient change from 0 to
centrifugation can be expressed as minus maximum gradient (-Gmax), after a delay of 5 times
1 r2 T 1, the other of half k-space data is collecting from '0' to
S= ∫
r2 − r1 r1
S (r )dr (8) '+kz ', corresponding to a gradient change from 0 to a
maximum gradient (Gmax). Fourier transformation of the k-
space data yields a real space image. In the centric scan
Equation (8) may be rewritten by changing the integration SPRITE method, the image signal intensity (S) is given by:
variable Pc(r2)=0 and Pc(r1)=PcL, with additional tp
mathematical manipulation, which yields the Hassler- S (r ) = M 0 (r ) exp(− ) sin α (12)

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

Brunner integral equation T2*
r1 + r2 S ( Pc ) where M0 is the equilibrium magnetization, α is the RF flip
∫
PcL
S PcL = dPc (9) angle, tp is the phase encoding time, T2* is the effective
2r2 0
Pc r2
1− (1 − 12 ) transverse relaxation time. Mo is directly proportional to the
PcL r2 local fluid content. Centric scan SPRITE methods are
naturally fluid content weighted.
Equation (9) cannot be directly solved for the unknown
function S. As pointed out by Hassler and Brunner [7], for a
short core, assuming r1/r2 ≈ 1, then equation (9) is reduced to
PcL
S PcL = ∫ S ( Pc )dPc (10)
0
whose differential form is
d
SL = ( S PcL ) (11)
dPcL

The value of PcL for each rotational speed is then calculated

with Eq. (7), and the value of saturation at inlet face, SL, is Figure 2. 1D centric scan SPRITE MRI method.
obtained according to Eq. (11). The relation of values of PcL
and SL, at different rotational speeds, yields the capillary Spin Density Imaging with SPRITE MRI
pressure curve. Equation (11) is an approximate solution A wide range of experimental results [16] show that the
introduced by Hassler and Brunner [7]. Based on equation (9), overall FID (free induction decay) decay rate (1/T2*) in
a number of other approximate solutions have been developed sedimentary rocks is dominated by an internal field
and used to determine capillary pressure curves [8, 18]. distribution (ΔBi) induced by the large susceptibility
difference (Δχ) between the pore fluid and solid matrix due to
SPRITE MRI paramagnetic impurities in the solid matrix. The decay rate of
the FID and the corresponding NMR linewidth (Δν=1/πT2*)
Standard SPRITE MRI for fluid saturated sedimentary rocks may be estimated by
The standard SPRITE MRI [14] technique has proven, over [16],
the last 9 years, to be a very robust and flexible method for the
study of a wide range of systems with short MR relaxation 1 γΔB i CΔχγB0
= Δν ≈ = (13)
times. As a pure phase encoding technique, SPRITE is largely πT2* 2π 2π
immune to image distortions due to susceptibility variation,
where γ is the gyromagnetic ratio, and B0 is the applied
chemical shift, and paramagnetic impurities. Repetitive
magnetic field strength, while C is a dimensionless constant.
excitation and data acquisition are performed in the presence
Equation (13) predicts a single exponential T2* decay,
of ramped phase encoding gradients, which enable systems
this prediction has been confirmed by a wide range of MR
with T2* lifetimes as short as tens of microseconds to be
experiments for sedimentary rocks [16]. Single exponential
successfully visualized.
T2* decay is anticipated for a wide variety of sedimentary rock
systems, but is not a universal result.
Centric scan SPRITE MRI
Figure 3 shows a semi-logarithmical FID decay after a
A centric scan strategy for SPRITE MRI [19] removes the
90 degree RF excitation pulse. The data was fit to the
longitudinal steady state from the image intensity equation of
equation:
standard SPRITE imaging, and increases the inherent image
S = M0 exp(-t/T2*) (14)
intensity. The image signal intensity no longer depends on the
where S is the MR signal intensity, t is the acquisition time.
longitudinal relaxation time and the repetition time. These
features ensure that centric scan SPRITE is an ideal method The fit T2* was 345 μs. We have observed for many
for quantitative imaging of sedimentary rocks with short sedimentary rocks that T2* is largely insensitive to water
relaxation times. saturation with a single exponential FID. These features
A 1D centric scan SPRITE technique is illustrated in ensure that Centric Scan SPRITE images are essentially spin
figure 2, where the k-space data are acquired sequentially density images. Spin density images may not be obtained by
4 SPE 100059

spin echo based MRI methods, due to multi-exponential T2 by increasing the image matrix size and/or decreasing the field
decay in rocks [11]. of view of the image, which results in more data points on the
capillary pressure curve.
100 Figures 4 and 5 show the 1D water content
FID signal (a.u.)

distribution along the length of the core before and after

centrifugation, respectively. The left side and right side of the
10 water content profile correspond to the inlet face and the outlet
face of the core, respectively. After centrifugation, the average
water saturation was 46.3%. Figure 6 shows the 1D water

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

1
0 500 1000
saturation distribution along the length of the core. The water
t (micro second)
saturation at the outlet face is 100%, which shows that the
outlet boundary condition is satisfied. Thus, MRI can be used
Figure 3. A semi-logarithmical plot of the MR FID (free induction to check the boundary condition for centrifuge capillary
decay) of the water saturated sandstone. The best fit line is single
exponential with a decay time constant T2* of 345μs.
pressure measurements.

SINGLE-SHOT CENTRIFUGE METHOD 1.0

A Berea sandstone and an oilfield sandstone samples were

Water content (a.u.)

employed for the capillary pressure curve measurements. The 0.8

dimensions of the core were measured to determine the bulk 0.6

volume of the rock sample. The samples were oven dried at 80
°C until a constant weight was reached, and the weight of the 0.4

dried sample was determined. The samples were kept under 0.2
vacuum conditions for more than 24 hours, and then saturated
with distilled and degassed water under vacuum conditions 0.0
0 2 4 6 8 10
until no bubbles came from the core samples. For the Berea Z (cm)
sandstone sample, the porosity was 0.19, and the permeability
was 0.20 Darcy. For the oilfield sandstone core #126, the Figure 4. One dimensional water content distribution along the
porosity was 0.28, and the permeability was 0.37 Darcy. Each length (Z) of the fully water saturated Berea sandstone core,
sample was 5.2 cm in length and 2.5 cm in diameter. determined with 1D centric scan SPRITE MRI.

Drainage Capillary Pressure Curve for Water and Air 1.0

The centrifuge experiment was carried out with a Beckman

Water content (a.u.)

0.8
TJ6R tabletop centrifuge with arm length of 18.6 cm, at 1500
RPM for 140 minutes. The temperature in the centrifuge was 0.6
controlled to be 3°C to avoid water evaporation during the
centrifuge experiment. The duration of centrifugation is 0.4

shorter than the duration typically used in crude/brine systems, 0.2

since the equilibrium time for water/air system is reduced. In
addition, the equilibrium time for high permeability rocks is 0.0

less than that for low-permeability rocks. A perforated Teflon 0 2 4 6 8 10

Z (cm)
end piece was employed to support the core and ensure full
water saturation at the outlet face. The cylindrical surface of
Figure 5. One dimensional water content distribution along the
the core sample was covered with a heat-shrink Teflon tube to length (Z) of the rock core after centrifugation, determined with 1D
maintain longitudinal fluid flow within the core. centric scan SPRITE MRI.
MR experiments were carried out in a 2.4 Tesla
horizontal bore superconducting magnet (Nalorac Cryogenics
100
Inc., Martinez, CA) with an Apollo console (Tecmag Inc.,
Water saturation (%)

Houston, TX). A proton-free 4.7 cm inner diameter eight-rung 80

quadrature birdcage probe (Morris Instruments, Ottawa, ON) 60

was employed. The core samples were wrapped with Teflon
40
tape to decrease the evaporation of water within the samples
during MR measurements. 20

The water content profiles along the length of the 0

core before and after centrifugation, were determined by 1D 0 1 2 3 4 5 6 7 8
centric scan SPRITE MRI with a phase encoding time of Z (cm)
30μs, flip angle α of 6 degrees, for a field of view of 9 cm,
with an image matrix size of 64 points. Four signal averages Figure 6. One dimensional water saturation distribution along the
were acquired for a total acquisition time of 25 seconds. More length (Z) of the Berea sandstone core after centrifugation,
determined with 1D centric scan SPRITE MRI. The water saturation
data points along the length of the core may easily be obtained is 100% at the outlet face of the core.
SPE 100059 5

25 Drainage and Imbibition Capillary Pressure Curves

for Water and Oil
20
Experimental procedures
Pc (psi)

15 In the single-shot measurement for capillary pressure curves

of water and oil, it is necessary to distinguish between oil and
10
water phases by MRI. This may be undertaken several
5
different ways, for example, employing D2O for the water
phase to make the water phase invisible to 1H MRI.

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

0 Experimental procedures for primary drainage, imbibition, and
0 20 40 60 80 100 secondary drainage capillary pressure curves follow:
Water saturation (%) 1. The cylindrical core sample was dried at 80°C until a
constant weight was reached, and the weight of the
Figure 7. A drainage capillary pressure curve of the Berea dried sample was determined.
sandstone core determined with the single-shot method. The 2. The core sample was kept under vacuum conditions
solid line is an interpolation through the experimental data points. for more than 24 hours, and then saturated with
The capillary pressure distribution can be calculated distilled water (H2O) under vacuum conditions until
with equation (6). Therefore, the relationship between no bubbles came from the core, and the weight of the
capillary pressure and the corresponding water saturation can saturated sample was determined.
be determined very straightforwardly. The capillary pressure 3. A proton (1H) one-dimensional Centric Scan SPRITE
curve is shown in Figure 7. MRI measurement was carried out on the sample
The capillary pressure curve obtained with water and which was wrapped with Teflon tape to decrease the
air can be converted to a capillary pressure curve under evaporation of liquid within the core sample during
mercury injection conditions according to the following MRI measurements. A water distribution along the
equation. length of the sample was obtained and normalized by
the total volume of water in the core sample.
PcHg (σ cos θ ) Hg
= (15) 4. Steps 1 to 2 were repeated, D2O instead of H2O was
Pcw− a
(σ cos θ ) w−a employed to saturate the core sample.
where PcHg and Pcw-a are the capillary pressures under mercury 5. The core sample was put into an oil filled sample tube
intrusion conditions and water-air conditions, respectively; for centrifugation at 3000 rpm in a centrifuge for 24
(σcosθ)Hg and (σcosθ)w-a are the products of the interfacial hours for primary drainage capillary pressure curve
tension and cosine of the contact angle under mercury measurement.
injection conditions and water-air conditions, respectively. 6. The core sample was taken from centrifuge, and its
weight determined.
For mercury injection, σ = 480 mN/m and θ = 140
7. A proton (1H) double half k-space SPRITE MRI
degree; for a water and air system, σ = 72 mN/m and θ = 0
measurement was carried out on the sample which
degree. A comparison of the capillary pressure curves
was wrapped with Teflon tape to decrease the
obtained by the proposed single-shot method and by mercury
evaporation liquid within the core sample during MRI
injection porosimetry is shown in Figure 8. The two methods
measurements. The oil distribution along the length of
yield remarkably consistent results.
the sample was obtained and normalized by the total
volume of oil in the sample. The D2O distribution
along the length of the sample was obtained by
100
subtracting the volume normalized oil distribution
80 from the volume normalized water distribution along
the length of the core sample. The water saturation
Pc (psi)

60 (Sw) distribution along the length of the core sample

after centrifugation was determined from the volume
40 normalized D2O distribution after centrifugation
divided by the volume normalized H2O distribution
20
before centrifugation. The capillary pressure curve
0
was determined with saturation distribution and
0 20 40 60 80 100 capillary pressure distributions along the length of the
Wetting phase saturation (%) core sample.
8. The core sample was flipped and spun in the
Figure 8. A comparison of drainage capillary pressure curves of centrifuge, then flipped and spun again several times
the Berea sandstone determined by mercury injection method (▲) to reach a relatively uniform irreducible water
and by a single-shot measurement (○) for water and air. The
capillary pressure curves were normalized with respect to surface saturation (Swi) distribution along the length of the
tension and contact angle. core sample to prepare the core sample for imbibition
capillary pressure curve measurement.
6 SPE 100059

9. The core sample was put into a D2O filled sample determined in practice. The current methodology shows they
tube for centrifugation at 3000 rpm in centrifuge for may now be determined with some ease.
48 hours for imbibition capillary pressure curve
measurement. 100
10. Steps 6 to 7 were repeated.
11. The core sample was flipped and spun in the 80

Sw (% )
centrifuge, then flipped and spun again several times 60
to reach a relative uniform residual oil saturation (Sor)
distribution along the length of the core sample to 40

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

prepare the core sample for secondary drainage 20
capillary pressure curve measurement.
0
12. The core sample was put into an oil filled sample tube
0 2 4 6 8 10
for centrifugation at 3000 rpm in the centrifuge for 48
Z (cm)
hours for the secondary drainage capillary pressure
curve measurement. Figure 9. Water saturation distribution along the length (Z) of core
13. Steps 6 to 7 were repeated again. #126 after centrifugation at 3000 rpm in oil for 24 hours for
primary drainage capillary pressure measurement. The water
saturation is 100% at the outlet face of the core.
Experimental results
A cylindrical sandstone sample #126 was employed for the
measurements. The dimensions of the core were measured to 100
determine the bulk volume of the rock sample. The capillary
pressure measurements were conducted according to the 80

Sw (% )
experimental procedures in the previous section. 60
The centrifuge experiment was carried out with a
Z513K tabletop centrifuge (Hermle Labortechnik, Germany) 40
with arm length of 23.5 cm, at 3000 RPM for 24 or 48 hours. 20
The MRI experiments were carried out in a 0.2 Tesla
0
permanent magnet with an Apollo console (Tecmag Inc.,
0 2 4 6 8 10
Houston, TX). A 3 cm inner diameter solenoid probe was
Z (cm)
employed. The core sample was wrapped with Teflon tape to
decrease the evaporation of liquid within the samples during Figure 10. Water saturation distribution along the length (Z) of
core #126 after centrifugation at 3000 rpm in water for 48 hours
MRI measurements. The advantages of using a low-field MRI for imbibition capillary pressure measurement.
instrument are that (1) the instrument is cheap, and (2) the
effective spin-spin relaxation time (T2*) is much longer than
100
the phase encoding time (tp). This ensures the SPRITE MRI
image is a simple fluid distribution image. 80
Sw (% )

In the experiments, decane was employed as the oil 60

phase. Crude oil can also be used in the experiment to
40
simulate reservoir conditions. The fluid content profiles along
the length of the core before and after centrifugation, were 20
determined by one-dimensional centric scan SPRITE MRI 0
with a phase encoding time of 50 μs, flip angle α of 6 degrees, 0 2 4 6 8 10
with an image matrix size of 64 points. 16 signal averages Z (cm)
were acquired for a total acquisition time of one minute. More Figure 11. Water saturation distribution along the length (Z) of
data points along the length of the core can easily be obtained core #126 after centrifugation at 3000 rpm in oil for 48 hours for
secondary drainage capillary pressure measurement.
by increasing the image matrix size and/or decreasing the field
of view of the image. This results in more data points on the 60
capillary pressure curve. 40
Figures 9, 10 and 11 show the 1D water saturation
distribution along the length of the core after primary 20
Pc (psi)

drainage, imbibition, and secondary drainage, respectively. 0

Therefore, the relationship between capillary pressure and the 0 20 40 60 80 100
corresponding water saturation can be determined very -20
straightforwardly. The primary drainage, imbibition, and -40
secondary drainage capillary pressure curves are shown in
Figure 12. The curves of Figure 12 are physically sensible and -60
closely approximate similar curves in the literature [20]. Sw(%)
However, the difficulty of conventional centrifuge Figure 12. Primary drainage (○), imbibition (□), and secondary
measurement methods means such curves are rarely drainage (Δ) capillary pressure curves obtained by the single-shot
method.
SPE 100059 7

DISCUSION gravity can also be neglected for the single-shot method. In

The single-shot method for determining the capillary pressure addition, the long rock cores preferred for the single-shot
curve is rapid, cheap, accurate and adaptable. The single method result in a relatively small radial effect [24] and
rotational speed for the single-shot measurement may be reduced rotational speed; some friable and unconsolidated
determined by the Leverett J function [21] with a J value of 4 samples may be applicable for the single-shot method, since
at the irreducible water saturation [22]. The capillary pressure extreme high rotational speeds are not necessary. The new
curve can be obtained straightforwardly with about 40 data centrifuge method can also be applied to inhomogeneous core
points. More data can be obtained simply by increasing the samples. The traditional method may not be able to identify
spatial resolution of the MRI experiment. A typical traditional the existence of heterogeneity of the sample, even though it

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

centrifuge for the capillary pressure measurement employs 10 requires a homogeneous core sample. These ideas will be
rotational speeds with 4 or 6 samples running simultaneously, elaborated in a future publication.
however, the new centrifuge method requires only a single
rotational speed with a maximum of 30 samples running CONCLUSIONS
simultaneously. The overall duration of the experiment for the The single-shot method combines a single-speed centrifuge
single-shot method is thus more than 50 times reduced from experiment and a quantitative SPRITE MRI technique to
the traditional centrifuge method. In other words, we believe directly determine the fluid saturation distribution along the
that the one single-shot system will have a sample throughput length of the core. The proposed method for determining
equivalent to 50 more ultracentrifuge systems theoretically. A capillary pressure curve is rapid, cheap, accurate and
cheap desktop centrifuge and a relatively inexpensive desktop adaptable. The capillary pressure curve can be obtained
permanent magnet based one-dimensional MRI instrument can straightforwardly with approximately 40 spatial image data
be employed for the single-shot method, instead of the very points. The duration of the experiment is much less than the
expensive traditional ultracentrifuge system. traditional centrifuge method. For the single-shot
In the single-shot method, the centrifugation measurement, the outlet boundary condition can be maintained
experiment is just for sample preparation, the MRI is by free water level, the effect of gravity can be neglected, and
employed for the capillary pressure curve measurement. The there is no effect of centrifuge speed variations on pore
MRI experiment for a fluid distribution measurement takes structure of rocks. In addition, a long rock core preferred for
less than 1 minute. Therefore, two or more desktop centrifuges the single-shot method result in a relatively small radial effect
may work with one MRI instrument to increase the capacity compared with a short sample; some friable and
dramatically without significant infrastructure cost increase. unconsolidated samples may be applicable for the single-shot
The overall duration of the new experiment is even faster than method. A comparison of the capillary pressure curves
mercury intrusion measurements for large quantities of obtained with the single-shot method and with the mercury
samples. injection method, yields remarkably consistent results. The
The test duration is very important for reservoir single-shot method has been applied to measure the primary
evaluation, especially for short term decision making drainage, imbibition, and secondary drainage capillary
processes. Since the traditional centrifuge and porous plate pressure curves for reservoir rocks.
capillary pressure curve measurements are slow, they often The single-shot method should make it possible for
can not assist the short term decision making processes for oil the capillary pressure curve measurement to become a
and gas exploration and production, even though capillary ‘routine’ rather than a ‘special’ core analysis measurement.
pressure is the fundamental property for multiphase flow in
the reservoir. ACKNOWLEDGEMENTS
The free water level maintained in the single-shot BJB thanks NSERC of Canada for discovery and equipment
method ensures the outlet boundary condition is maintained. grants. BJB also thanks the Canada Chairs program for a
The single-shot method can measure the primary drainage, as Research Chair in MRI of Materials (2002-2009). The UNB
well as positive and negative imbibition and secondary MRI Centre is supported by an NSERC Major Facilities
drainage capillary pressure curves for reservoir rocks by Access grant. The authors thank the Idea to Innovation
maintaining the free water level within the cores. The program of NSERC for funding support and R.P. MacGregor
traditional centrifuge method however can only measure the for his technical assistance.
negative imbibition curve and positive secondary drainage
curve. Therefore, it can not be used to determine the Amott REFERENCES
wettability index. After primary drainage and imbibition [1] Lake L., Enhanced Oil Recovery, Prentice Hall, New Jersey,
measurements for the traditional and new centrifuge methods, (1989), 48.
the rock core should be managed to reach a uniform [2] Purcell W., Capillary pressure – Their measurement using
irreducible water saturation distribution and uniform residual mercury and the calculation of permeability therefrom, Trans. AIME
186, 39 (1949).
oil saturation distribution, respectively. MRI was also [3] Fatt I. and H. Dyksta, Relative permeability studies, Trans AIME
employed to check the saturation distribution, but traditional 192, 41 (1951).
methods can not measure and check the fluid distribution. [4] Burdine N., Relative permeability calculations from pore size
For the single-shot method, since only a single distribution data, Trans. AIME 198, 71 (1953).
rotational speed is employed, there is no pressure effect on [5] Masalmeh S., X. Jing, W. van Vark, S. Christiansen, H. van der
rock pore structure [23] due to dramatic rational speed Weerd and J. van Dorp, Impact of SCAL on carbonate reservoirs:
changes required for the traditional method. The effect of how capillary forces can affect field performance predictions, SCA
8 SPE 100059

paper 2003-36 in Proceedings of the International Symposium of the

Society of Core Analysts, Pau, France, Sept. 21-24, (2003).
[6] Dullien F., Porous Media: fluid transport and pore structure, 2nd
Edition, Academic Press, New York (1991), 139-176.
[7] Hassler G. and E. Brunner, Measurement of capillary pressures in
small core samples, Trans. AIME, 160, 114 (1945).
[8] Ruth D. and Z. Chen, On the measurement and interpretation of
centrifuge capillary pressure curves: The SCA survey data, The Log
Analyst 36, 21 (1995).
[9] Melrose J., Interpretation of Centrifuge Capillary Pressure Data,

Downloaded from http://onepetro.org/SPEIOR/proceedings-pdf/06IOR/All-06IOR/SPE-100059-MS/1826584/spe-100059-ms.pdf/1 by U. of Alberta Library user on 15 October 2021

paper N in SPWLA 27th Annual Logging Symposium Transactions,
Houston, Texas, June 9-13 (1986).
[10] Rajan R., Theoretically Correct Analytical Solution for
Calculating Capillary Pressure-Saturation from Centrifuge
Experiments, paper J in SPWLA 27th Annual Logging Symposium
Transactions, Houston, Texas, June 9-13 (1986).
[11] Baldwin B. and E. Spinler, A direct method for simultaneously
determining positive and negative capillary pressure curves in
reservoir rock, J. Pet. Sci. Eng. 20, 161 (1998).
[12] Graue A.,T. Bognø, E. Spinler and B. Baldwin, A method for
measuring in-situ capillary pressures at different wettabilities using
live crude oil at reservoir conditions, part 1: feasibility study, SCA
paper 2002-18 in Proceedings of the International Symposium of the
Society of Core Analysts, Monterey, California, USA, 22-25 Sept.
(2002).
[13] Chen Q. and B. Balcom, Measurement of rock-core capillary
pressure curves using a single-speed centrifuge and one-dimensional
magnetic-resonance imaging, J. Chem. Phys. 122, 214720 (2005).
[14] Balcom B., J. Barrita, C. Choi, S. Beyea, D. Goodyear and T.
Bremner, Single-point magnetic resonance imaging (MRI) of cement
based materials, Materials and Structures 36, 166 (2003).
[15] Balcom B., R. MacGregor, S. Beyea, D. Green, R. Armstrong
and T. Bremner, Single Point Ramped Imaging with T1 Enhancement
(SPRITE), J. Magn. Reson. A 123, 131 (1996).
[16] Chen Q., A. Marble, B. Colpitts, and B. Balcom, The internal
magnetic field distribution, and single exponential magnetic
resonance free induction decay, in rocks, J. Magn. Reson. 175, 300
(2005).
[17] Bear J., Dynamics of Fluids in Porous Media, Dover
Publications, New York, (1972), 444.
[18] Forbes P., Centrifuge data analysis techniques, an SCA survey
on the calculation of drainage capillary pressure curves from
centrifuge measurements, SCA paper 9714 in Proceedings of the
International Symposium of the Society of Core Analysts, Calgary,
Sept. 8-10, (1997).
[19] Mastikhin I., B. Balcom, P. Prado and C. Kennedy, SPRITE
MRI with Prepared Magnetization and Centric k Space Sampling, J.
Magn. Reson. 136, 159 (1999).
[20] Donaldson E. C., Thomas R. D. and Lorenz P. B.,Wettability
determination and its effect on recovery efficiency, SPE Journal,
9(1), 13 (1969).
[21] Leverett M., Capillary behaviour in porous solids, Trans. AIME,
142, 152 (1941).
[22] Brown H. W., Capillary pressure investigations, Trans. AIME,
192, 67 (1951).
[23] Chen Q., W. Kinzelbach, C. Ye and Y. Yue, Variations of
permeability and pore size distribution of porous media with
pressure, J. of Environmental Quality, 31, 500 (2002).
[24] Christansen R. and K. Cerise, Geometric concerns for accurate
measurement of capillary pressure relationships with centrifuge
methods, SPE Formation Evaluation, Dec. 311 (1988).