Characterizing Disproportionate

Permeability Reduction Using Synchrotron

X-Ray Computed Microtomography
R.S. Seright, SPE, New Mexico Petroleum Recovery Research Center; J. Liang, SPE, Idaho Natl. Engineering and
Environmental Laboratory; W. Brent Lindquist, SPE, State U. of New York at Stony Brook; and John H. Dunsmuir,
ExxonMobil Research & Engineering Co.

Summary (CCD). Using Fourier methods, the set of angular projections at

X-ray computed microtomography was used to investigate why each row of pixels in the CCD was used to reconstruct the cross-
gels reduce permeability to water more than that to oil in strongly sectional slice at that row. These slices were stacked to form the
water-wet Berea sandstone and in an oil-wet porous polyethylene 3D image. In this work, a 5× microscope objective was used to
core. Although the two porous media had very different porosities provide a pixel size of 4.1 ␮m and a 4.1-mm field of view. Because
(22% vs. 40%), the distributions of pore sizes and aspect ratios part of the core was outside the imaged area, a profile extension
were similar. A Cr(III)-acetate-HPAM gel caused comparable oil method was used to supress edge artifacts.
and water permeability reductions in both porous media. In both Several authors used XMT to characterize the microscopic
cores, the gel reduced permeability to water by a factor 80 to 90 structure of porous media.17,19,23 For a 15-darcy sandstone, Coles
times more than that to oil. However, the distributions of water and et al.19 found a mean tortuosity of 2.7, with a range from 1.5 to 4.5.
oil saturations (vs. pore size) were substantially different before, Along a 2.2-mm-long section of this core, porosity varied only a
during, and after gel placement. This paper examines the mecha- few percent around the average value (26.4%). After oilflooding,
nism for the disproportionate permeability reduction in the two this core was waterflooded to a water saturation of 25.1%. Inter-
porous media. estingly, large variations in water saturation were observed along
the 2.2-mm-long section, ranging from 12 to 39%. A 3D view
Introduction showed the nonwetting phase (water, in this case) to exist as large
Many polymers and gels can reduce the permeability to water ganglia (blobs of nonwetting phase that extend over multiple
more than that to oil or gas.1–15 This property is critical to the pores, often exhibiting a branched structure).19
success of water-shutoff treatments in production wells if hydro- Chatzis et al.24,25 suggested that rock heterogeneity can be
carbon-productive zones cannot be protected during polymer or responsible for saturation variations within a porous medium. Non-
gelant placement.2,3 However, the magnitude of the effect has been wetting phase saturations that are lower than expected can occur
unpredictable from one application to the next. Presumably, the when clusters of small pores are dispersed in a matrix dominated
effect would be more predictable and controllable if we understood by large pores. In contrast, nonwetting phase saturations that are
why the phenomenon occurs. Although many mechanisms have higher than expected can occur when clusters of large pores are
been considered (see Table 1), the underlying cause of the dis- dispersed in a matrix dominated by small pores.24 However, sig-
proportionate permeability reduction remains elusive. nificant saturation variations can occur even in homogeneous po-
Previously, we used NMR imaging to observe disproportionate rous media, depending on the pore-body/pore-throat aspect ratio.
permeability reduction on a microscopic scale.16 Results from For homogeneous 2D micromodels, Chatzis et al.24 reported pis-
these experiments revealed that the imaging technique had many tonlike displacements with very little trapping of the nonwetting
limitations that prevented us from obtaining reliable pore-level phase when the aspect ratio was 2 or less. However, for aspect
images. Most importantly, the spatial resolution was on the order ratios around 3, large nonwetting phase clusters formed as the
of hundreds of micrometers, which was too low to clearly distin- wetting phase formed fingers while displacing the nonwetting
guish fluid pathways on the pore level. phase. At higher aspect ratios, the nonwetting phase tended to be
In this paper, we describe imaging experiments using high- trapped in individual pores rather than in large clusters of pores.
resolution computed X-ray microtomography (XMT) to compare The pore coordination number had a minor effect on nonwetting
the oil and water pathways and fluid distributions before and after phase residual saturations.24
gel treatment. The current generation of synchrotron-based XMT Using XMT data, Lindquist et al.23 extensively characterized
scanners provides the ability to obtain 3D pore-level images of pore- and throat-size distributions for Fontainebleau sandstones.
rock samples with a spatial resolution on the order of microme- As core porosity increased from 7.5 to 22%, they found that the
ters.17–23 For this study, we used the ExxonMobil beamline X2-B average pore coordination number increased from 3.4 to 3.8; the
at the Natl. Synchrotron Light Source.18 X2-B is a dedicated XMT average channel length decreased from 200 to 130 ␮m; the average
imaging facility capable of producing continuous registered stacks throat area increased from 1,600 to 2,200 ␮m2; and the average
of 2,048 × 2,048 × 1,024 14-bit 3D images of X-ray linear attenu- pore volume remained fairly constant at approximately 0.0004
ation coefficients at energies tunable from 8 to 40 keV. The highly mm3. The aspect ratio (effective average pore radius/effective av-
collimated synchrotron X-rays permit the reconstruction of a 3D erage throat radius) was greater than 2 in 65% of pores and greater
image from 2D projections taken at uniformly spaced angles be- than 3 in 40%. The aspect ratios tended to increase slightly as
tween 0 and 180°. X2-B converts the pattern X-rays transmitted by porosity decreased.
the specimen (projections) to a visible light image with a thin
single crystal of CsI(Na). This image was magnified by an optical Experimental Materials
microscope objective onto a 1,024 × 1,024 charge coupled device We performed two sets of imaging experiments with strongly wa-
ter-wet Berea sandstone cores and one set with a polyethylene core
that was strongly oil-wet. (The water-advancing contact angle was
measured at 165° for the brine/oil/polyethylene system used in this
Copyright © 2002 Society of Petroleum Engineers
work.) The Berea cores had permeabilities of ∼0.47 darcies and
This paper (SPE 79717) was revised for publication from paper SPE 71508, first presented porosities of 22%. The polyethylene core had a permeability of 8.8
at the 2001 SPE Annual Technical Conference and Exhibition, New Orleans, 30 Septem-
ber–3 October. Original manuscript received for review 18 November 2001. Revised manu-
darcies and a porosity of 40%. (Consistent porosity values were
script received 9 July 2002. Paper peer approved 30 July 2002. determined from both image analyses and conventional mass-

October 2002 SPE Reservoir Evaluation & Engineering 355

 Fig. 1—Image cross section of Berea core saturated with brine.
balance measurements.) The cores were 6.5 mm in diameter and 1.15 mm x 1.15 mm.
30 mm in length, with an intermediate pressure tap 6 mm from the
inlet face. Our scans focused on a segment of the core that was 6.5
mm in diameter and 3.25 mm in length. To minimize end effects, defined as the smallest cross-sectional area of a channel.) Pore
the scanned segment was located approximately halfway between bodies are regions delimited by throats.
the inlet and outlet faces. The brine used during the waterfloods Fluid phase location is detected by subtracting X-ray attenua-
contained 1% NaCl and 0.1% CaCl2. A hexadecane mixture was tion coefficients in a two-fluid-phase filled image against the cor-
used during the oilfloods. To increase the image contrast between responding image containing only one of the fluid phases in the
the brine and oil phases, hexadecane was doped with 10% w/w void space. (Precision alignment between the two images is re-
iodohexadecane in our first set of experiments in Berea and with quired.) As grain/void space identification can be performed on the
15% w/w bromohexadecane in the second set. For the polyethyl- single fluid-phase image, this (voxel-by-voxel) subtraction is done
ene core, the hexadecane was doped with 15% w/w bromohexa- only for the fluid-filled pore voxels.
decane. All experiments occurred at room temperature. The gelant To visualize the pore structure and fluid locations, we first
used in these experiments contained 0.5% Alcoflood 935 HPAM focused on thin cross sections with dimensions of 1.15 mm × 1.15
(molecular weight ≈5 × 106 daltons; degree of hydrolysis ⳱ 5 to mm in the x-y directions (i.e., perpendicular to the flow direction).
10%), 0.0417% Cr(III) acetate, 1% NaCl, and 0.1% CaCl2. The Figs. 1 and 2 illustrate image cross sections for the first Berea core
gelant viscosity at room temperature was 20 cp. The viscosities when saturated with brine and for the polyethylene core when
were 1.0 cp for brine (without polymer), 3.3 cp for the hexadecane/
iodohexadecane mixture, and 3.6 cp for the hexadecane/
bromohexadecane mixture.

Core Characterization
Three-dimensional scans were performed after a gelant flood and
after oil and waterfloods both before and after gel placement.
(Images were acquired at saturation endpoints.) All floods in a
given set of experiments were conducted without removing the
core from the sample stand so that the images could be compared
directly. For each scan the image was cropped into a 2.97 mm ×
2.97 mm × 2.1 mm rectangular block (the images were 725 × 725
× 512 voxels at 4.1 ␮m/voxel) to remove artifacts caused by those
parts of the core that did not remain within the field of view
through all 180° of sample rotation. Image analyses were per-
formed with a software package called 3DMA—a statistical analy-
sis tool that correlates saturations with geometry. This software is
capable of measuring distributions of pore size, pore-body/pore-
throat aspect ratio, and coordination number from 3D XMT images
of porous rock. Specific details on the algorithms used by this
software can be found in Ref. 23. Briefly, the software first ana-
lyzes an image of a rock containing a single fluid phase in the pore
space. A segmentation algorithm that uses the X-ray attenuation
coefficient information in the image assigns each cubic volume
element (voxel) as either grain or void. A medial axis algorithm is
used to trace the percolating backbone (skeleton) of the void space.
Properly interpreted, coordination numbers and channel lengths
can be obtained directly from the backbone. The backbone is also Fig. 2—Image cross section of polyethylene core saturated with
used as a search device in locating channel throats. (A throat is oil. 1.15 mm x 1.15 mm.

356 October 2002 SPE Reservoir Evaluation & Engineering

Fig. 3—Scanning electron micrograph of Berea. 30 µm x 22 µm. Fig. 4—Scanning electron micrograph of polyethylene. 30 µm x
22 µm.
saturated with oil. The black areas show rock grains or polyeth-
ylene. The white areas show brine-saturated voids in the Berea and in pores with a given size (indicated on the x-axis). Again, the
oil-saturated voids in the polyethylene. These figures highlight the distributions were quite similar, considering the material differ-
irregular sizes and shapes of the voids. Although the porosity ences. The peak in the pore volume occurred at an effective pore
difference is noticeable (22% for Berea and 40% for polyethylene), radius (assuming spherical pores): approximately 50 ␮m for the
the pore body and throat sizes were comparable. Electron micro- first Berea core, 70 ␮m for the second Berea core, and 70 ␮m for
graphs (Figs. 3 and 4) were obtained at high magnification for the polyethylene core. The average nodal pore volumes for the
Berea and polyethylene to illustrate the character of the pore walls. three cores were very similar to those found by Lindquist et al.23
(These images were obtained with a Cameca SX100TM electron for Fontainebleau sandstones (∼0.0004 mm3). A comparison of
microprobe at New Mexico Tech.) Because of a coating of kaolin- Figs. 5 and 6 reveals that although many pores existed with vol-
ite, a significantly greater surface roughness and angularity existed umes less than 0.0001 mm3, their contribution to the total void
in Berea (Fig. 3) than in the polyethylene core (Fig. 4). Interest- volume was small.
ingly, before exposure to oil, thin filaments (∼0.1-␮m diameter)
bridged many of the small pores in polyethylene. However, these Aspect Ratios. The distribution of aspect ratios (effective pore
filaments were destroyed upon exposure to hexadecane and were radius/effective throat radius) for the Berea and polyethylene cores
found to be not relevant to our results.26 is shown in Fig. 7. (The effective pore radius computed is for a
sphere with a volume equivalent to that measured for the pore. The
Size Distributions. The pore-size distributions for two Berea effective throat radius computed is for a circle with an area equiva-
cores and the polyethylene core are shown in Fig. 5. The y-axis lent to that measured for the throat.) The y-axis plots the average
plots the percent of the total number of pores that had a given pore aspect ratio in pores with a given size (indicated on the x-axis).
volume (indicated on the x-axis). The distributions (based on pore Again, the distributions were surprisingly similar for the cores.
numbers) were similar for the three cores. For pores with volumes The average aspect ratio was 4.0 for the first Berea core, 4.2 for the
below 0.0003 mm3, the fraction of pores of a given size was fairly second Berea core, and 4.4 for the polyethylene core. As pore
insensitive to the pore volume. Above 0.0003 mm3, the concentration volume increased from 10–5 mm3 (effective pore radius ∼13 ␮m)
of pores decreased significantly with increased pore volume. to 0.002 mm3 (effective pore radius ∼78 ␮m), the average aspect
Pore-volume distributions for the three cores are shown in Fig. ratio increased steadily from 2 to 6. Aspect ratios jumped sharply
6. The y-axis plots the percent of the total void volume that existed for the few largest pores. For a given pore size, a wide range of
aspect ratios were noted. For all cores at a given pore size, the
100
Percent of Pores With Given Pore Volume

20
1st Berea core
2nd Berea core
Percent of Total Volume

10
Polyethylene core
15

1
10

1st Berea core

0.1 2nd Berea core
5
Polyethylene core

0.01 0
1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01
Pore Volume, mm3 Pore Volume, mm3

Fig. 5—Pore-size distributions. Fig. 6—Pore-volume distributions.

October 2002 SPE Reservoir Evaluation & Engineering 357

 40
16 1st Berea 0.0001<PV<0.001 mm 3
core 0.001<PV<0.01 mm 3
30

% of Total
Polyethylene
12 PV>0.01 mm 3
core 20
Aspect Ratio

2nd Berea
core
10
8
PV<0.0001 mm 3
0

<3

<5
<2

0
0

0
4

<1

>3
<2

<3
R

R
R

R
<R

<R
2<

3<

5<

10

20
Aspect Ratio, R
0
1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 Fig. 8—Aspect ratios in Berea.
Pore Volume, mm3

Fig. 7—Aspect-ratio distributions. number of exits from a pore.) The y-axis plots the average coor-
dination number in pores with a given size (indicated on the x-
axis). The average coordination number was 3.9 for the first Berea
standard deviation (of aspect ratios) was typically 65% of the
core, 4.7 for the second Berea core, and 6.2 for the polyethylene
mean value.
core. For the smallest pores, the coordination number was approxi-
For Berea sandstone, Fig. 8 plots the distributions of aspect
mately 3 for all three cores. As the pore size increased, the coor-
ratios for each of four ranges of pore volume (PV): (1) PV>0.01
dination numbers increased, with the polyethylene core experienc-
mm3; (2) 0.001<PV<0.01 mm3; (3) 0.0001<PV<0.001 mm3; and
(4) PV<0.0001 mm3. The x-axis plots various ranges of aspect
ratio, R, from R<2 up to R>30. The y-axis plots the percent of the
total aspect ratios (for a given PV range) that fall within a given
range of aspect ratios. The solid and open triangles in Fig. 8 show
that approximately 35% of the smallest pores were associated with
aspect ratios that were less than 2. Interestingly, for all four pore-
size ranges, 25 to 35% of the aspect ratios fell between 3 and 5,
and a significant percentage of aspect ratios fell between 5 and 10.
Aspect ratios above 10 were common for the larger pores but were
rare for the smaller pores. In contrast, aspect ratios below 3 were
very common for the smaller pores but were much less frequent for
the larger pores.
The average throat area was 1,330 ␮m2 for the first Berea core,
1,460 ␮m2 for the second Berea core, and 1,630 ␮m2 for the
polyethylene core. These values were generally lower than the
average throat areas reported for Fontainebleau sandstones (1,600
to 2,200 ␮m2).23

Coordination Numbers. The distributions of pore coordination

numbers are shown in Fig. 9. (The coordination number is the

100
2nd Berea
core
Coordination Number

Polyethylene
core
10
1st Berea
core

1
1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01
Pore Volume, mm3
Fig. 10—Berea sandstone cross-sectional image slices (1.15
Fig. 9—Coordination number distributions. mm × 1.15 mm each). Green is water. Red is oil. Black is rock.

358 October 2002 SPE Reservoir Evaluation & Engineering

ing a slightly more rapid increase than the Berea cores. Coordina-
tion numbers up to 70 were noted for the largest pores. For a given
pore size, standard deviations were typically 20 to 40% of the
mean values.
Ioannidis et al.27 reported comparable properties for Berea
sandstone based on analyses of photomicrographs of 78 serial
sections through a double pore cast of a Berea sample. Their
analyses of 1,268 pores and 1,945 throats indicated that the aver-
age coordination number was 3.46—somewhat lower than our
values. Their average throat area was 60 to 90% greater than our
values, and their pore-size distribution was weighted toward larger
pores than our distributions.

Images After the Various Floods

XMT scans were performed after each flood. Fig. 1 and Fig. 10
show image slices for the first Berea core for each of the flooding
stages. Figs. 2 and 11 show image slices for the various flooding
stages in the polyethylene core. In Figs. 10 and 11, oil is red, water
is green, and the rock grains (or polyethylene) are black. To mag-
nify the pore structure, these cross sections are 1.15 mm × 1.15
mm. Corresponding fluid saturations and permeabilities associated tion method, which was used to distinguish between oil and water
with the floods are listed in Tables 2 and 3. For several cases in phases. Recall that fluid-phase identification is made by subtract-
Figs. 10 and 11, small isolated oil drops or thin oil filaments (in ing two corresponding images: one with a single phase in the void
red) appear floating in water (in green) within a single pore. In space, the other having both fluid phases present (one phase of
other cases, isolated water drops or thin water filaments appear which is doped with a strong X-ray attenuating compound). A
floating in oil within a single pore. We recognize that multiple histogram of the subtracted X-ray attenuation coefficients will
droplets and suspended liquid filaments are unlikely in individual show a bimodal distribution, with the peaks corresponding, respec-
pores. The chief origin of these artifacts is our current segmenta- tively, to water- and oil-filled voxels. The peaks in this histogram
will overlap. There are several reasons for the overlap: the finite
size of a voxel means that any X-ray attenuation coefficient mea-
sured is an average over the voxel volume; variation in X-ray count-
ing statistics results in variations in attenuation coefficients; and mi-
nor alignment errors, on the order of 1⁄2–1 voxel width, do occur.
Currently, we use simple thresholding to distinguish between
water- and oil-filled voxels. The threshold is determined from an
examination of the bimodal subtraction histogram, with the thresh-
old value picked to lie in the “valley” between the two peaks.
Simple thresholding is well known to result in speculated images
(i.e., apparent blobs of one phase trapped in the other). The extent
of the speculation is proportional to the overlap between the two
peaks in the difference histogram. Examination of our data indi-
cates that such speculation accounts for less than 6% of the volume
of any one phase. Note additionally that caution must be observed
in evaluating the amount of speculation by viewing only 2D slices
such as Figs. 10 and 11. Often, what appears as a disconnected
blob in two dimensions is a piece of a shaped surface of a much
larger 3D structure. In fact, this can be noticed in Fig. 1, which
appears to show a few disconnected grain segments. Analysis of
the full 3D image indicates no disconnected grains.

Berea Sandstone Image Analyses

Before Gelant Injection. In the first Berea core, the first scan
viewed the core with 100% brine saturation (Fig. 10a). Second, the

Fig. 11—Polyethylene cross-sectional image slices (1.15 mm ×

1.15 mm each). Green is water. Red is oil. Black is polyethylene.

October 2002 SPE Reservoir Evaluation & Engineering 359

core was flooded with ∼35 core pore volumes of oil and scanned As mentioned, the overall average Sor was 43.5% for the first
at residual water saturation (Fig. 10b). Oil invaded most pores, Berea core. This number was significantly higher than the Sor
while the residual water appeared dominantly in crevices and films values measured with mass balances in this strongly water-wet
between the oil and the rock. The residual water saturation, Swr, Berea sandstone (∼22%). At Sor, the average water saturation was
was 24.7% (Table 2), and the endpoint relative permeability to oil surprisingly insensitive to pore size (open circles in Fig. 12). These
(relative to the absolute permeability), kro, was 0.82 (Table 3). results were confirmed during imaging studies of our second core
Third, the core was flooded with water (∼70 core pore volumes) (open squares in Fig. 12).
and scanned at residual oil saturation (Fig. 10c). Pores appear with For a given pore size, Fig. 12 shows the average water satura-
a range of saturations (i.e., individual pores are not necessarily tion from a given experiment. Of course, because many pores were
nearly all oil-filled or nearly all water-filled). The residual oil present for any given pore size, a range of saturations was found.
saturation, Sor, was 43.5% (i.e., 100%–56.5%, from Table 2), and A detailed examination of the data revealed that for a given pore
the endpoint relative permeability to water, krw, was 0.16. These size, the range or distribution of water saturations was quite broad,
findings are consistent with expectations in a strongly water-wet rock. both at Sor and at Swr. One feature of these XMT studies is that we
Using the 3DMA software, detailed analyses were performed can monitor the saturation differences for individual pores from
with the complete 3D images (not just the 2D slices shown in Figs. one flooding stage to the next. Because of space limitations, this
10 and 11). The distributions of water and oil saturations were level of detail is not reported here but can be found in Ref. 26. For
determined as a function of pore size. In Fig. 12, the solid symbols example, at Sor, virtually all pores contained at least 20% oil, and
show the distribution of water saturations (Sw) at Swr before gel half the pores contained at least 45% oil. Chatzis et al.24 noted that
for pore-body/pore-throat radius aspect ratios around 3, the non-
placement. (In this figure and subsequent figures, average satura-
wetting phase tends to be trapped in large clusters of pores. In
tions for a given pore size are reported.) The open symbols show
contrast, for ratios exceeding 3, the nonwetting phase tends to be
the distribution of water saturations at Sor before gel placement.
trapped in individual pores. We noted that 75% of our pores had
(Of course, at any condition, the oil saturation is equal to 100%
aspect ratios exceeding 2, and 56% had ratios exceeding 3.
minus the water saturation.) As mentioned, the overall average Swr
While oil saturations at values intermediate between 0 and 1
was 24.7% for the first Berea core. This number was consistent
would be unexpected in a medium containing only spherical pores,
with the Swr values measured using mass balances in this strongly
the images in Figs. 1 through 4 and Figs. 10 and 11 show that the
water-wet Berea sandstone. As expected at Swr, the water satura-
pore shapes are typically elongated and asymmetric, containing
tion generally increased with decreased pore size (solid circles in
corners, nooks, and crannies. The combination of individual pore
Fig. 12). A broad minimum in water saturation was noted around
trapping and irregular pore-body shapes allows for a wide range of
0.002 mm3, and a local maximum was observed around 0.01 mm3.
intermediate saturations for individual pores.
For a strongly water-wet porous medium, one might have expected
Because our cores were small (30 mm in length), the capillary
the water saturation to approach 100% for the smallest pores.
end effect could have created a saturation gradient through the
Instead, the water saturation in the smallest pores averaged 60%. A
core. Three factors worked in our favor to minimize this effect.
calculation using the Young-Laplace equation confirmed that oil
First, the image volume was located near the center of the core,
should be able to enter the small pores (i.e., PV ∼10–5mm3) in our
maximizing distance from the inlet and outlet faces. Second, our
Berea cores. Specifically, to enter the smallest pore throats (∼7
cores were quite permeable (0.47 darcies for Berea and 8.8 darcies
␮m), a capillary pressure around 1 psi was needed. This value was
for polyethylene). Third, our floods were conducted with relatively
much lower than the 17 psi (pressure drop across the core) that was
high capillary numbers and pressure gradients (17 to 35 psi/ft).
applied during our flooding experiments. The behavior of Sw vs.
Examination of our images (which cover approximately 7% of the
pore size was confirmed during an imaging experiment in a second
core length) revealed no obvious saturation trend in any direction.
Berea core (solid squares in Fig. 12). This experiment included a
Nevertheless, we cannot discount the possibility that a saturation
second oilflood (following an intervening waterflood). Results
gradient might have existed in the cores.
from this part of the experiment (solid triangles in Fig. 12) further
One also might wonder whether our small image volume (∼19
confirmed the above behavior and indicated limited hysteresis dur-
mm3) was less than the representative elementary volume (REV)
ing cycles of water- and oilflooding before gel placement. The
for this porous medium. We are unable to answer this question
fluid saturations and relative permeabilities from this second Berea
definitively. Our Berea images contained 1,700 to 1,800 pores that
core (Tables 2 and 3) also suggest that little hysteresis occurred
were larger than 10–5 mm3. Those who advocate the REV concept
during multiple flooding cycles.
might argue that at least 1 million pores (i.e., 100 grain diameters
per side of the sample volume) are needed to form an REV. Ob-
viously, our image volume does not meet this criterion. On the
100 other hand, the observations that we report in this paper appear to
1st core, at 1st Swr
be generally valid throughout the imaged volumes (i.e., we did not
2nd core, at 1st Swr observe regions of exceptional behavior within a given image vol-
80 2nd core, at 2nd Swr ume). This point can be appreciated in part by considering Figs. 10
Water Saturation, %

1st core, at Sor and 11.

2nd core, at Sor
60 During Gelant Injection. After establishing a residual oil satura-
tion in the first Berea core, the 20-cp Cr(III)-acetate-HPAM gelant
(10 core pore volumes) was injected with a pressure gradient of 17
40 psi/ft. After gel placement, the core was shut in for 12 hours (at
60°C) and then scanned (at room temperature). The gelation time
for this formulation was 1 to 1.5 hours at 60°C.
20 A comparison of Figs. 10c and 10d suggests that fluid satura-
tions changed somewhat during gelant injection. A detailed analy-
sis of the 3D images (Fig. 13) confirmed that fluid saturations
changed during gelant injection. The open and solid circles in Fig.
0
13 reveal that average water saturations decreased slightly in the
0.00001 0.0001 0.001 0.01 0.1 medium-to-large pores during gelant injection in Berea. Also, the
Pore Volume, mm3 overall average water saturation decreased from 56.5 to 47.7%
(Table 2). On first consideration, these observations surprisingly
Fig. 12—Water saturations at Swr and Sor before gel placement suggest that oil was generated when gelant (containing no oil) was
in Berea sandstone cores. injected. However, the reader should recall that the imaged volume

360 October 2002 SPE Reservoir Evaluation & Engineering

 100 was incomplete) or (2) that oilflooding after gel placement
moved, concentrated, or destroyed much of the gel that formed in
the oil pathways.
80 Finally, brine (2.5 core pore volumes) was injected to measure
Water Saturation, % the water residual resistance factor, Frrw. The Frrw value was
1,220, meaning that at Sor, the gel reduced the permeability to
60 water by a factor of 1,220. This gel reduced the permeability to
water 81 times more than that to oil (Frrw/Frro⳱81). Fig. 10f
shows the core scan after brine injection. A comparison of Figs.
40 10c and 10f shows that water did not have access to most of the
pathways that were open to oil flow after the gel treatment.
Berea before gelant The open symbols in Fig. 14 compare distributions of water
20
Berea after gelant saturations at Sor before and after gel placement. For reasons yet to
Polyethylene before gelant be explained, water saturations in the smallest pores (<0.0005
mm3) at Sor after gel placement were less than those at Swr. How-
Polyethylene after gelant ever, more importantly (because the total pore volume was domi-
0
nated by the largest pores), the distribution of water saturations in
0.00001 0.0001 0.001 0.01 0.1 the largest pores (>0.001 mm3) after gel placement during water-
Pore Volume, mm3 flooding (open triangles) was similar to that during the previous
oilfloods (solid symbols in Figs. 12 and 14). Therefore, the gel
Fig. 13—Water saturations at Sor immediately before vs. imme- treatment apparently trapped substantial additional volumes of oil
diately after gel placement. during waterflooding (i.e., much higher Sor values). Perhaps this
result occurred because gel was strategically positioned in pore
constituted only 10% of the total core volume. Conceivably, oil throats to increase aspect ratios. With most pores permanently
was displaced from upstream portions of the core during gelant occupied by oil, water was forced to flow through thin films,
injection, and this oil became trapped in the imaged volume by through the smallest pores, and through the gel itself, explaining
coincidence. Presumably, the overall oil content of the core either the large water residual resistance factor (i.e., 1,220). In contrast,
stayed the same or decreased slightly during gelant injection. oil pathways remained relatively free from constriction by the gel,
so the oil residual resistance factor was much less (i.e., 15).
After Gel Placement. After gel placement, oil (∼20 core pore Polyethylene Image Analyses
volumes) was injected to measure the oil residual resistance factor,
Frro, recording a value of 15. This value means that at Swr, the gel Before Gelant Injection. In the polyethylene core, the first scan
reduced the permeability to oil by a factor of 15. The core was viewed the core with 100% oil saturation (Fig. 2). Second, the core
scanned after oil injection to visualize oil pathways after gel (Fig. was waterflooded (∼70 core pore volumes) and scanned at residual
10e). A comparison of Figs. 10b and 10e indicates that most of the oil saturation (Fig. 11a). Water invaded most pores, while the
pathways open to oil flow before gel placement remained acces- residual oil appeared dominantly in small pores and films between
sible to oil after the gel treatment. This result suggests that after oil water and the polyethylene. The oil films were less evident than
injection, the gel occupied only a small fraction of the pore space. the water films in Berea (compare Figs. 10b and 11a); however,
Tracer results from a previous experiment in a high-permeability careful examination reveals that the oil films were present on the
Berea core revealed that gel with an Frro value of 20 occupied less polyethylene. In reality, the liquid films should be much thinner
than 5% of the pore space.1 than we can resolve with the XMT method. The appearance that
The solid symbols in Fig. 14 compare distributions of water water films on Berea sandstone were thicker than oil films on
saturations at Swr before and after gel placement. The two distri- polyethylene may be caused by the greater surface roughness in
butions were remarkably similar, confirming that most of the path- Berea (compare Figs. 3 and 4). Greater surface roughness creates
ways that were open to oil flow before gel placement were also folds of the film and fluid accumulations in nooks and crannies.
open to oil flow after gel placement. As mentioned earlier, this Although the actual film thickness may be only a few tens of
suggestion is consistent with the relatively low oil residual resis- nanometers at a given point, the film folds and small fluid accu-
tance factor (i.e., 15). Because the gel was placed and formed at mulations are interpreted to be much wider with our imaging tech-
high water saturation in the core, the result also suggests either (1) nique because the resolution is limited to the micrometer scale.
that gel did not form in all the aqueous pore space (i.e., gelation Third, oil (∼35 core pore volumes) was injected to drive the
core to residual water saturation (Fig. 11b). Residual water was
noted in pores with a wide range of sizes. Also, water often ap-
100 peared at the polyethylene interface, suggesting either very thin oil
at Swr before gel films or partial water wetting of the plastic.
at Sor before gel Water (∼70 core pore volumes) was again injected to establish
80 Sor (Fig. 11c). The similarity of Figs. 11a and 11c indicates little
Water Saturation, %

at Swr after gel

hysteresis during the cycle of water and oil injection. The similar-
at Sor after gel ity of water saturations (77.7% vs. 75.0%) and relative permeabili-
60 ties (0.50 vs. 0.49) for the two waterfloods at Sor supports this
suggestion (Tables 2 and 3). Fig. 15 provides additional support.
This figure plots water saturations as a function of pore size before
40
gelant injection. The solid circles show the distribution of water
saturations at Swr before gel placement. The open circles and tri-
angles show the two distributions at Sor. The similarity of the two
20
distributions at Sor confirms that hysteresis was not significant.
A comparison of Figs. 12 and 15 reveals significant differences
0
in the saturation distributions for Berea vs. polyethylene. Most
notably, the intermediate-to-large pores exhibited a substantially
0.00001 0.0001 0.001 0.01 0.1
greater saturation difference between Swr and Sor for polyethylene
Pore Volume, mm3 (∼15% vs. ∼85%) than for Berea (∼30% vs. ∼60%). For the small-
est pores in both porous media, the distributions at Swr converged
Fig. 14—Effect of gel on Swr and Sor in Berea sandstone. with those at Sor. However, the water saturation for this conver-

October 2002 SPE Reservoir Evaluation & Engineering 361

 100 number, viscoelastic polymer solutions (e.g., our gelant) redistrib-
ute forces on a microscopic scale in a manner that drives residual
phases to lower-than-expected values. These possibilities will be
80 explored during future work.
Water Saturation, %
at 1st Sor After Gel Placement. After the shut-in period, oil (20 core pore
at Swr volumes) was injected, and a value of 24 was measured for the oil
60 at 2nd Sor residual resistance factor. This value was quite similar to the cor-
responding value measured in Berea (15, from Table 3). However,
a comparison of the XMT scans (Figs. 10e and 11e) suggests that
40 the oil saturations and oil pathways were very different in poly-
ethylene from those in Berea. In Berea, the overall oil saturation
was quite high (71%, from Table 2). In polyethylene, the oil satu-
20 ration was much lower (24.3%, from Table 2). Interestingly, the oil
locations in Fig. 11e were very similar to those in Figs. 11a and
11c. Thus, in polyethylene, the oil locations during oil injection
after gel placement were virtually the same as those during water
0 injection before gel placement. This observation receives further
0.00001 0.0001 0.001 0.01 0.1 support from the plots of water saturation vs. pore size. In Fig. 16,
Pore Volume, mm3 the saturation distribution at Sor before gel placement (open
circles) was virtually the same as that at Swr after gel placement
Fig. 15—Water saturations at Swr and Sor before gel placement
(solid triangles).
in the polyethylene core. Finally, water (2.5 core pore volumes) was injected, and a value
of 2,130 was measured for Frrw. This value was comparable to that
for Berea (1,220, from Table 3). Also, the ratio Frrw /Frro was 89
gence occurred at a significantly lower value for polyethylene (10 in polyethylene vs. 81 in Berea. Despite the similarity of Frrw
to 20%) than for Berea (50 to 60%). For polyethylene, the average values, the water saturations and water pathways appear very dif-
water saturation at Swr was nearly independent of pore size. In ferent in polyethylene from those in Berea (compare Figs. 10f and
contrast, in Berea, the average water saturation at Sor was nearly 11f). In Berea, a comparison of Figs. 10e and 10f suggests that the
independent of pore size. For polyethylene at Sor, the water satu- dramatic permeability reduction for water was caused by gel trap-
ration increased significantly with increased pore size. In contrast, ping a high residual oil saturation. In contrast, in polyethylene
for Berea at Swr, the water saturation decreased significantly with (Fig. 11f), the oil saturation was very low (7.6%, from Table 2).
increased pore size. Presumably, these differences reflect the pref- Interestingly, the images shown in Figs. 11d and 11f are very
erence for water to occupy the smallest pores in water-wet Berea similar. This observation receives further support from the plots of
and not to occupy the smallest pores in the oil-wet polyethylene water saturation vs. pore size. The saturation distribution at Sor
(i.e., because the ratio of surface area to volume increased with immediately after gel placement (solid triangles in Fig. 13) was
decreased pore size). very similar to that after the final waterflood after gel placement
(open triangles in Fig. 16).
During Gelant Injection. In the next step, gelant (10 core pore Presumably during gelant injection, virtually all water was dis-
volumes) was injected with a pressure gradient of 23 psi/ft. After placed by the aqueous gelant. As mentioned earlier, some residual
gel placement, the core was shut in for 14 hours (at ∼70°C) and oil also was displaced. After gelation, an effective gel permanently
scanned (at room temperature). A comparison of Figs. 11c and 11d occupied locations that were previously occupied by water at Sor
indicates that oil was mobilized during gelant injection. Immedi- before gelant injection. Thus, oil could no longer flow through
ately after gel placement, most remaining oil appeared as a film those locations. The only flow paths available for oil were those
attached to the polyethylene. Although a water residual resistance through locations that were occupied by oil at Sor before gelant
factor was not measured at this point, in a separate study,13 an Frrw injection. This scenario explains why the distribution of saturations
value of 25,000 was measured for this gelant and core material at Sor before gelant injection was the same as that at Swr after gel
immediately after gel placement. This high residual resistance fac-
tor indicated that the gel occupied virtually all of the aqueous pore 100
space and reduced the core permeability to approximately 0.2 md.
A detailed analysis of the 3D images (Fig. 13) confirmed that
fluid saturations changed during gelant injection. The open and
solid triangles in Fig. 13 reveal that average water saturations 80
Water Saturation, %

increased substantially in the small-to-medium pores during gelant at Swr before gel
injection in polyethylene. Also, the overall average water satura-
tion increased from 75.0 to 94.4% (Table 2). On the one hand, an at Sor before gel
60
increase in water saturation might be expected because the injected at Swr after gel
gelant was 20 times more viscous than the displaced water. How-
ever, on the other hand, the pressure gradient during gelant place- at Sor after gel
ment (23 psi/ft) was intentionally kept below that used during the 40
other flooding steps (35 psi/ft) to avoid mobilization of the residual
phase. Three arguments could be offered to rationalize the ob-
served behavior. First, because polyethylene was oil-wet, a con-
tinuous oil film may have existed throughout the porous medium. 20
Thus, true irreducible oil saturation may not exist. With increased
gelant or water throughput, oil in the continuous film may be able
to drain slowly from the porous medium. A second explanation is 0
that the gelant may have changed the wettability of the porous
0.00001 0.0001 0.001 0.01 0.1
medium to more water-wet. However, although some argue in
favor of this type of mechanism,28 it is counterintuitive and re- Pore Volume, mm3
quires much more convincing support. A third possible mechanism
was suggested by Wang et al.29 Specifically, at a fixed capillary Fig. 16—Effect of gel on Swr and Sor in polyethylene.

362 October 2002 SPE Reservoir Evaluation & Engineering

placement (Figs. 11c, 11e, and 16). However, at the condition Conclusions
associated with Sor, why should oil be able to flow at all? Perhaps X-ray computed microtomography was used to investigate why
a continuous oil film on the polyethylene allowed flow (i.e., the oil gels reduce permeability to water more than that to oil in strongly
was not trapped as discontinuous drops at Sor). Considering that water-wet Berea sandstone and in an oil-wet porous polyethylene
Frro was 24 for this experiment, the above suggestion implies that core. The following conclusions were reached:
if no gel was present, the permeability to oil at a water saturation 1. Although the two porous media had very different porosities
of 75.7% (from Table 2) should be 24 times less than the perme- (22% vs. 40%), the distributions of pore sizes and aspect ratios
ability to oil at a water saturation of 15.2%. In other words, per- were similar.
haps our core was not at true irreducible oil saturation when water 2. A Cr(III)-acetate-HPAM gel caused comparable oil and water
was injected. We will explore this possibility in future work. permeability reductions in both porous media. In both cores, the
Recall that the process of gel placement drove Sor from 25.0 to gel reduced permeability to water by a factor 80 to 90 times
5.6% (Table 2). However, during oilflooding after gel placement, more than that to oil.
oil returned to those locations where gelant had previously re- 3. The distributions of water and oil saturations (vs. pore size)
moved residual oil (see Figs. 11c through 11e and Figs. 13 and 16). were substantially different before, during, and after gel
Why did gel not block these locations permanently? Considering placement.
the size of these locations (red areas of Figs. 11c through 11e) and 4. The disproportionate permeability reduction appeared to occur
the inert nature of polyethylene, it seems likely that gel formed by different mechanisms in the two porous media. In Berea, gel
here as effectively as in other locations. Possibly, during the oil- caused disproportionate permeability reduction by trapping sub-
flood after gel placement, the topography of the porous medium, stantial volumes of oil that remained immobile during water-
combined with capillary forces, allowed the oil to rip9,10 or dehy- flooding. With this high trapped-oil saturation, water was forced
drate15 gel to form pathways in the regions occupied by the origi- to flow through thin films, through the smallest pores, and
nal oil saturation. As mentioned earlier, this mechanism also may through the gel itself. In contrast, during oilflooding, oil path-
have been operable in Berea. However, in Berea, the gel trapped ways remained relatively free from constriction by the gel.
oil associated with saturations at Swr before gel placement. In 5. In the polyethylene core, oil trapping did not contribute signifi-
polyethylene, the gel trapped oil associated with saturations at Sor cantly to the disproportionate permeability reduction. Instead,
before gel placement. oil films and a relatively small number of pore pathways pro-
After gel placement, the permeability to water was dramatically vided conduits for the oil. For reasons yet to be understood, the
less than that to oil. This result suggests that the largest oil path- small pore pathways appeared largely unavailable for
water flow.
ways shown in Figs. 11c and 11e generally were not open to water
flow. Why not? One proposed possibility was that a new residual
Nomenclature
oil saturation was established in these pathways during the final
waterflood.15 However, careful examination of Figs. 11c through Frr ⳱ residual resistance factor
11f and Figs. 13 and 16 reveals that this did not happen. A second Frro ⳱ residual resistance factor for oil
possibility is that elastic forces associated with the gel closed the Frrw ⳱ residual resistance factor for water
largest oil pathways during the final waterflood.9,10 In this theory, kro ⳱ relative permeability to oil
capillary forces keep the oil pathways open during oilflooding.9,10 krw ⳱ relative permeability to water
A third possibility is that the oil pathways were open during oil- R ⳱ aspect ratio
flooding but closed during waterflooding because gels “shrink in Sor ⳱ residual oil saturation, %
the presence of oil but swell in the presence of water.”4–6 This Sw ⳱ water saturation, %
proposed mechanism is troubling because in the absence of os- Swr ⳱ residual water saturation, %
motic effects, the available evidence indicates that our gels do not
shrink in the presence of oil and swell in the presence of water.7,9
Acknowledgments
Of course, additional work will be performed to address the
questions and test the hypotheses raised in this paper. Financial support for this work is gratefully acknowledged from
the Natl. Petroleum Technology Office (NETL) of the U.S. Dept.
of Energy, BP, Chevron, China Natl. Petroleum Corp., Chinese
Relevance to Field Applications Petroleum Corp., Marathon, Shell, and Texaco. The efforts of John
Hagstrom were appreciated during the corefloods and imaging
When production wells were treated with gelants, some zones had
experiments. We also thank Jill S. Buckley (New Mexico Tech)
high water saturations and high fractional water flows, while other
for helpful discussions and suggestions and Nelia Dunbar (New
zones had high hydrocarbon saturations and fractional flows. For
Mexico Tech) for obtaining the electron micrographs. This re-
the oil zones, the flooding sequence used for this paper mimics the
search was carried out (in part) at the Natl. Synchrotron Light
field situation. Specifically, oil was the first fluid to flow through
Source, Brookhaven Natl. Laboratory, which is supported by the
the gel after the treatment. Thus, the oil residual resistance factor
U.S. Dept. of Energy, Division of Materials Sciences and Division
of 15 and the mechanism for establishment of oil permeability
of Chemical Sciences. The Geosciences Program of the U.S. Dept.
through the gel are relevant to field applications.1–3 However, this
of Energy (grant DE-FG02-92ER14261) funded development of
is not the case for the water zones. Specifically, in water zones in
the 3DMA code.
field treatments, water was the first fluid to flow through the gel
after gel placement (i.e., no oil flowed through the gel in the water
References
zones). For this circumstance, our previous experience with
Cr(III)-acetate-HPAM gels revealed that water residual resistance 1. Liang, J., Sun, H., and Seright, R.S.: “Reduction of Oil and Water
factors were ∼10,000 and that the gel occupied virtually all of the Permeabilities Using Gels,” paper SPE 24195 presented at the 1992
aqueous pore space.1,2 Thus, the mechanism for water permeabil- SPE/DOE Symposium on Enhanced Oil Recovery, Tulsa, 22–24 April.
ity reduction in field applications is actually one of near-total 2. Seright, R.S., Liang, J., and Sun, H.: “Gel Treatments in Production
blockage of water flow. Virtually the only means for water flow is Wells with Water Coning Problems,” In Situ (1993) 17, No. 3, 243.
through the gel itself (which has a permeability less than 1 3. Liang, J., Lee, R.L., and Seright, R.S.: “Gel Placement in Production
md).16,27 In contrast, in the experiments here, the mechanism of Wells,” SPEPF (November 1993) 276; Trans., AIME, 295.
water flow was examined after oil flowed through the gel. Nev- 4. Sparlin, D.D. and Hagen, R.W. Jr.: “Controlling Water in Producing
ertheless, because the water residual resistance factor in these ex- Operation—Part 5,” World Oil (July 1984) 137.
periments was 80 to 90 times greater than the oil residual resis- 5. Dawe, R.A. and Zhang, Y.: “Mechanistic Study of the Selective Action
tance factor, the mechanism of permeability reduction for water is of Oil and Water Penetrating into a Gel Emplaced in a Porous Me-
of scientific interest. dium,” J. Pet. Sci. Eng. (1994) 12, 113.

October 2002 SPE Reservoir Evaluation & Engineering 363

 6. Mennella, A. et al.: “Pore-Scale Mechanism for Selective Permeability Intl. Symposium on Evaluation of Reservoir Wettability and its Effect
Reduction by Polymer Injection,” paper SPE 39634 presented at the on Oil Recovery, Montpellier, France, 11–13 September.
1998 SPE/DOE Improved Oil Recovery Symposium, Tulsa, 19–22 23. Lindquist, W.B. et al.: “Pore and Throat Size Distributions Measured
April. From Synchrotron X-Ray Tomography Images of Fountainebleau
7. Liang, J., Sun, H., and Seright, R.S.: “Why Do Gels Reduce Water Sandstones,” J. Geophys. Research (2000) 105B, 21508.
Permeability More Than Oil Permeability?” SPERE (November 1995) 24. Chatzis, I., Morrow, N.R., and Lim, H.T.: “Magnitude and Detailed
282; Trans., AIME, 299. Structure of Residual Oil Saturation,” SPEJ (April 1983) 311.
8. Zaitoun, A., Rahbari, R., and Kohler N.: “Thin Polyacrylamide Gels for 25. Chatzis, I., Kuntamukkula, M.S., and Morrow, N.R.: “Effect of Cap-
Water Control in High-Permeability Production Wells,” paper SPE illary Number on the Microstructure of Residual Oil in Strongly Water-
22785 presented at the 1991 SPE Annual Technical Conference and Wet Sandstones,” SPERE (August 1988) 902.
Exhibition, Dallas, 6–9 October. 26. Seright, R.S.: “Using Chemicals to Optimize Conformance Control in
9. Liang, J. and Seright, R.S.: “Further Investigations of Why Gels Re- Fractured Reservoirs,” annual report (DOE/BC/15110-6), Contract No.
duce Water Permeability More Than Oil Permeability,” SPEPF (No- DE-AC26-98BC15110, U.S. DOE (2001) 2–39.
vember 1997) 225. 27. Ioannidis, M.A. et al.: “Comprehensive Pore Structure Characteriza-
10. Al-Sharji, H.H. et al.: “Pore-Scale Study of the Flow of Oil and Water tion Using 3D Computer Reconstruction and Stochastic Modeling,”
Through Polymer Gels,” paper SPE 56738 presented at the 1999 SPE paper SPE 38713 presented at the 1997 SPE Annual Technical Con-
Annual Technical Conference and Exhibition, Houston, 3–6 October. ference and Exhibition, San Antonio, Texas, 5–8 October.
11. Zaitoun, A., Bertin, H., and Lasseux, D.: “Two-Phase Flow Property 28. Elmkies, P. et al.: “Further Investigations on Two-Phase Flow Property
Modifications by Polymer Adsorption,” paper SPE 39631 presented at Modification by Polymers: Wettability Effects,” paper SPE 64986 pre-
the 1998 SPE/DOE Improved Oil Recovery Symposium, Tulsa, 19–22 sented at the 2001 SPE International Symposium on Oilfield Chemis-
April. try, Houston, 13–16 February.
12. Nilsson, S., Stavland, A., and Jonsbraten, H.C.: “Mechanistic Study of 29. Wang, D. et al.: “Study of the Mechanism of Polymer Solution With
Disproportionate Permeability Reduction,” paper SPE 39635 presented Visco-Elastic Behavior Increasing Microscopic Oil Displacement Ef-
at the 1998 SPE/DOE Improved Oil Recovery Symposium, Tulsa, ficiency and the Forming of Steady ‘Oil Thread’ Flow Channels,”
19–22 April. paper SPE 68723 presented at the 2001 SPE Asia Pacific Oil and Gas
13. Liang, J. and Seright, R.S.: “Wall-Effect/Gel-Droplet Model of Dis- Conference and Exhibition, Jakarta, 17–19 April.
proportionate Permeability Reduction,” SPEJ (September 2001) 268.
14. White, J.L., Goddard, J.E., and Phillips, H.M.: “Use of Polymers To
Control Water Production in Oil Wells,” JPT (February 1973) 143. SI Metric Conversion Factors
15. Willhite, G.P. et al.: “Mechanisms Causing Disproportionate Perme- bbl × 1.589 873 E–01 ⳱ m3
ability in Porous Media Treated With Chromium Acetate/HPAAM cp × 1.0* E–03 ⳱ Pa·s
Gels,” paper SPE 59345 presented at the 2000 SPE/DOE Improved Oil eV × 1.602 19 E–19 ⳱ J
Recovery Symposium, Tulsa, 3–5 April.
ft × 3.048* E–01 ⳱ m
16. Seright, R.S.: “Improved Techniques for Fluid Diversion In Oil Re-
covery,” final report, DOE/BC/14880-15, U.S. DOE (January 1996) °F (°F– 32)/1.8 ⳱ °C
97–108. in. × 2.54* E+00 ⳱ cm
17. Dunsmuir, J.H. et al.: “X-Ray Microtomography: A New Tool for the psi × 6.894 757 E+00 ⳱ kPa
Characterization of Porous Media,” paper SPE 22860 presented at the *Conversion factor is exact.
1991 SPE Annual Technical Conference and Exhibition, Dallas, 6–9
October.
18. Flannery, B.P. et al.: “Three-Dimensional X-Ray Microtomography,” Randy Seright is a senior engineer at the New Mexico Petro-
Science (18 September 1987) 237, 1439. leum Recovery Research Center at New Mexico Tech in So-
corro. e-mail: [email protected]. He holds a PhD degree in
19. Coles, M.E. et al.: “Developments in Synchrotron X-Ray Microtomog-
chemical engineering from the U. of Wisconsin at Madison. He
raphy With Applications to Flow in Porous Media,” SPEREE (August has served on the SPE Board of Directors and the Board Com-
1998) 288. mittee on Publications. Jenn-Tai Liang is an advisory engineer/
20. Zhou, M. et al.: “Irreducible Water Distribution in Sandstone Rock: scientist at the Idaho Natl. Engineering and Environmental
Two Phase Flow Simulations in CT-based Pore Network,” Phys. Chem. Laboratory. e-mail: [email protected]. Liang holds a PhD degree
Earth (A) (2000) 25, No. 2, 169. in petroleum engineering from the U. of Texas at Austin. He is
21. Coles, M.E. et al.: “Pore Level Imaging of Fluid Transport Using a member of the Editorial Review Committee. Brent Lindquist
Synchrotron X-Ray Microtomography,” paper 9628 presented at the is a professor in the Dept. of Applied Mathematics and Statis-
tics at the State U. of New York at Stony Brook. e-mail:
1996 Intl. Symposium of the Soc. of Core Analysts, Montpellier,
[email protected]. He holds a PhD degree in physics
France, 8–10 September. from Cornell. John Dunsmuir is a member of the technical staff
22. Hazlett, R.D., Chen, S.Y., and Soll, W.E.: “Wettability and Rate Effects in the Advanced Materials Characterization Section at
on Immiscible Displacement: Lattice Boltzmann Simulation in Micro- Exxon Mobil Research and Engineering Corporate Strate-
tomographic Images of Reservoir Rocks,” paper presented at the 1996 gic Research.

364 October 2002 SPE Reservoir Evaluation & Engineering