state-of-the-art review of nondestructive testing with computer-assited tomography
DESCRIPTION
State-of-the-Art Review of Nondestructive Testing With Computer-Assited TomographyTRANSCRIPT
SPESociety of Petroleum Engineers
SPE 22127
State-of-the-Art Review of Nondestructive Testing WithComputer-Assisted TomographyA. Bansal, U. of Alaska Fairbanks, and M.R. Islam, EMERTEC Developments Inc.SPE Members
Copyright 1991, Society of Petroleum Engineers, Inc.
This paper was prepared for presentation at the International Arctic Technology Conference held in Anchorage, Alaska, May 29-31, 1991.
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 presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessanly reflectany position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Societyof Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgmentof where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A. Telex, 730989 SPEDAL.
ABSTRACT
Computer assisted tomography (CAT) has been a
revolutionary technique in medical radiology.
Recently, CAT scanners started to be used as
non-destructive testing facilities in different
industrial applications. Most of the applications
of the new technology are in the areas of petroleum
engineering. Majority of this CAT scanner
technology has been developed in the U.S.A.
However, several Canadian companies have acquired
CAT scanners and are using for novel commercial as
well as research applications. This paper presentsa comprehensive review of advances made in the
areas of CAT scan technology as applied to thepetroleum industry.
Recent efforts in developing CAT scan technology
has been in the areas of core analysis methods.
The ultimate objective is the quantitative
monitoring of dynamic two- and three-phase
experiments in porous media. Even though advances
have been made in using CAT scan for quantitative
evaluation of fluid flow, more research has to be
done before gaining confidence m such a technique.
This paper presents a critical review of recent
research publications. Also, discussed is the
References at end of paper
523
possibility of using CAT scan as a unique reservoir
characterization tool. Reservoir properties, such
as direction and frequency, of fractures, thepresence of vuggs, stratification in the bedding
plane, etc. may be identified only through
non-destructive testing of reservoir cores.
Various applications of CAT scanning to oil
industry are discussed in detail.
PRINCIPLES OF CAT
Recently, Kantzas1 has outlined in detail the
principle of computer assisted tomography.
Consequently, we will describe the principle only
briefly. The original theory for reconstructing a
complete image of an object from numerous views
around the object was first described by Radon in
19172• The principle of CAT is based on emitting
x-ray from a source which revolves around the
object in consideration while one-dimensional
projections of attenuated x-rays are collected by a
detector on the other side of the source. These
projections are collected as the sample travels
through the scanner longitudinally and are used to
reconstruct a three-dimensional image of the
object. Intensity values of attenuated x-rays are
collected from small volumetric elements, called
pixels. These elements are typically O.76mmxO.76mm
2STATE-OF-THE-ART REVIEW OF NON-DESTRUCTIVE TESTING
WITH COMPUTER-ASSISTED TOMOGRAPHY SPE 22127
2n 00
f(r,lP) = #(r,lP) = J J F(R,O) e2niRrcos(IP-0)IRldRdO
o 0
where 1 is the distance along the scan line, ° is
the angle describing the scan line, R is the
conjugate Fourier frequency variable and p(l,O) is
the transmission of a particular ray described by
(1,0). The distribution is then found by
transforming back into real space, given by:
where M(r, IP) is the linear attenuation coefficient
given as a function of r and IP. The variable r is
obtained from 1 by r = 1 cos( IP-O) and arises from
the Jacobian of a coordinate transformation. The
angle IP represents a point in real space and the
angle ° represents a point in Fourier space. The
above two equations are sufficient for
reconstructing a two-dimensional object from its
Here dl IdE represents the spectral distribution ofo
intensity, e(E) the efficiency of the detector at a
given energy E, el the lower energy and eh the
higher energy. Also, # is now a function of:
spatial coordinates as well as photon energy. One:
can see immediately that the volume of data as well
the complexity of data processing increases
drastically by implementing this relationship. As
a consequence, this relationship is not commonly
used. In stead, an effective energy is assumed to
replace a spectrum of energy and the resulting
error is corrected by manipulating the data for
beam hardening effects. The resulting problem,
thus, involves in determining two-dimensional
distribution of p(x,y) from the line integrals.
Several methods are available for solving this
problem. One such method is the Fourier method
derived from the Central Projection Theorem. This
theorem states that the Fourier transforms of a
one-dimensional projection of a two-dimensional
distribution are mathematically, identical to a
section through a two-dimensional Fourier transform
of the original distribution. The Fourier
transform of the transmission profiles is given in
polar coordinates as
+00
J p(1,O) e-2niRl d1
-00
F(R,O)
eh d
I = J(dI/dE) e(E) exp [ - J#(x,y,E) ds ] dE
e l s
where #* is the linear attenuation coefficient, L
is the path length, I is the incident intensityo
and I the transmitted intensity. Of course, #* is
characteristics of the medium and reflects the
nature of the medium. This relationship applies
only for a narrow monoenergetic beam of x-ray
photons which travels across a homogeneous medium.
If the medium in consideration is heterogeneous,
the above equation holds true while replaced by the
line integral of the linear attenuation
coefficients. The modified form is:
d
In(l/I) = J#(x,y)dL
s
,/ = (lIL) In(III)o
in area and 1 cm in depth (along the direction of
the x-ray beam) for a second generation CAT
scanner. Once these elements are all assigned an
intensity values after a complete radial and
longitudinal scan, these data are processed by a
computer. This processing constitutes the major
part of the CAT. The inlet intensity and the
outlet intensity are related through the following
relationship,
where dL is the differential of path length along
the beam, s is the x-ray source, and d is the
detector. As compared to the equation, a unique
value of # is replaced by a distribution #(x,y) in
plane perpendicular to the direction of the x-ray
beam. This relationship calls for a large memory
for storing all the information even for a
relatively small sample. Besides, this
relationship is still quite idealized as it assumes
the x-ray beam to be monoenergetic. In reality,
the beam includes a wide range of photons ranging
from 20 keY up to 120 keY in some instances. Also,
the detector system in itself is energy dependent.
Consequently, a rigorous representation of the
intensity relationship is given by
524
SPE 22127 A. Bansal and M.R. Islam 3
one-dimensional projections. Before this
conversion can be used, the Fourier coefficients
need to be converted from polar to rectangular grid
coordinates and the two-dimensional Fourier
transform needs to be taken to obtain the density
distribution on a Cartesian grid which is usually
used for computer display systems.
Another method, called the Convolution Method or
Back Projection, is more commonly used for
commercial CAT systems. This method is
mathematically close to the Fourier method. In
this method, the process is represented by
where p is the projection value and LI Ok is the
angular increment between projections. The above
relationship is equivalent to throwing back the
projection data for each of the various views into
the reconstruction plane and summing them for
obtaining an image. Obviously, this is a very
crude state of reconstruction and is known to bethe distorted version of the correct solution.
This distortion is removed by the two-dimensional
convolution of the degraded image with an
appropriate correlation function or by taking the
two-dimensional Fourier transform of the image
while multiplying by an appropriate filter and
performing the inverse Fourier transform. These
methods require considerably large computer time
and are known to give less than optimal results
with real data. This calls for an alternative
method, called the Filtered Back-Projection method.
This method uses correction and filtering of data
prior to back projecting. The filtering is done as
a convolution operation in the spatial domain or as
a functional multiplication in the Fourier
frequency domain. The Fourier filtered back
projection method multiplies the Fourier transform
of the projection data by a ramp-shaped filter
which has a magnitude proportional to the Fourier
frequency variable associated with the linear
position of the projection data, 1. The
convolution theorem dictates that this spatial
I frequency filtering is equivalent to convoluting
the projection data with a corresponding filter
525
function in the spatial domain. The corrected
projection data, p. (r.,O) is obtained from the1
measured projection data, p(r.,O), for any given1
projection angle 0, by the following relationship
p·(r.,O) = (p(r.,O)/4a) - (1I1r2a)[Z p(r.,O)/(i-j)2]1 1 J J
while counting only odd values of (i-j). In this
.relationship, a is the sampling interval. The
image is formed by back projecting these corrected
projection data onto a suitable two-dimensional
matrix by transforming back into real space as
shown by an earlier equation. The result of the
reconstruction process is, thus, a two-dimensionalarray of numbers. Several methods may be used for
representing the reconstruction matrix. These
matrix values are directly related to the linear
attenuation coefficient distribution of the object
at a given effective x-ray energy. Before storing
this information into a computer, the linear
attenuation coefficients are scaled to integer
values covering a given range and commonly adjusted
so that the value of water is zero. The following
equation relates the linear attenuation
coefficients to the number stored in computer
(known as the CT numbers or CTn)
where p and p are the linear attenuationwater
coefficients of the object scanned and of water at
the effective energy of the beam, and K is a
scaling factor equal to the number of CT values·
between air and water. Most present scanners use a
value of 1000 for K. Therefore, CT numbers range
from -1000 to +1000 with water corresponding to a
CT number of O. Also, each CT number corresponds
to 0.1 % of the linear attenuation coefficient of
water.
DETERMINATION OF PHYSICAL PROPERTIES
WITH CATSCAN
One of the most successful application of CAT
scanning lies in determining physical properties of
rocks non-destructively. The CAT Scanner produces
two-dimensional arrays of data that are called CT
numbers. In order to correlate the CT numbers to
4STATE-OF-THE-ART REVIEW OF NON-DESTRUCTIVE TESTING
WITH COMPUTER-ASSISTED TOMOGRAPHY SPE 22127
oillwater system, saturations are given by
S = A/Co
S 1 - Sg 1
For a liquid gas system, the following equations
provide estimation of saturations:
cP(CT -CT )o w
S = 1 - Sw 0
(CT -CT ) - cP(CT -CT )mew 3S =
o
For a three-phase system, saturations are given by
the following equations:
CTm-CT1
ij) (CT - CT)1 •
where subscripts a,w,0 denote air, water, and oil,
respectively, whereas m and e represent saturated
and dry state of the core, respectively. CT refers
to CT numbers of corresponding fluids (as given by
the subscripts).
Z3 0 8
JJ = p (a + P E3 0 2)
where a, p are constants, E is the mean photon
energy in keV, p is the bulk density and Z is the
effective atomic number of the sample. For
energies above 100 keY, x-rays interact with
materials more vigorously by Compton scattering
which is dependent on bulk density of the sample.
For x-ray energies below 100 keY, photoelectric
absorption becomes a dominant phenomenon and
depends on effective atomic number of each pixel.
Therefore, a dual energy system is used for usual
CAT scanning. The high energy scan can be used for
determining bulk density while the second scan at
low energy can be used for determining effective
atomic numbers. At this point, the porosity of the
pixel is given by
specific physical properties, one has to seek
relationship between linear attenuation
coefficients and other physical properties. Vinegar
and Wellington3 provided a detailed discussion on
this technique. They showed that the linear
attenuation coefficient is a function of the bulk
density and the effective atomic number of the
sample, given by
S = B/Cw
where, Pb is the bulk density, Pg is the mineral or
grain density, and Pf is the fluid density. Two
scans of the same porous medium would provide one
with bulk density, grain density (mineralogy) as
well as porosity. This would, in turn, provide one
with information on heterogeneities in a porous
medium.
S =1-S-Sg 0 W
where A, B, and C are defined by:
A = (CT -CT) (CT -CT) - (CT -CT) (CT -CT)m el W g2 m e2 w gl
B = (CT -CT) (CT -CT) - (CT -CT )I(CT -CT)2m e2 0 gl mew ~
Fluid saturation in a dynamic or static system can
be measured with CAT scanning as well. However,
typically scanning period has to be faster than
movement of fluid flow in a porous medium. This
could be a severe limitation if one wishes to study
the effect of high flow rates in immiscible or
miscible flood experiments. For a two-phase
C = cP«CT -CT ) (CT -CT) - (CT -CT )2(CT -CT )1)o gl w g2 0 g w g
Subscripts 1 and 2 refer to the scan at energy
levels of 1 and 2, respectively. These equations
are somewhat different from those of Wellington and
Vinegar4 in the sense that CT numbers are used of
those of fluids rather than cores saturated with
respective fluids. Kantzas~ as shown these two
formulations to be equivalent to each other.
526
SPE 22127 A. Bansal and M.R. Islam 5
APPLICATION OF CAT IN OIL INDUSTRY
Multiphase Flow Visualization and
Relative Permeability Measurements
Wang et aI.6 are reportedly the first researchers
to publish on CAT application of multiphase fluid
flow in porous media. They reproduced the image of
saturation distribution during an immiscible
liquid-liquid displacement in a Berea sandstone
core7
• The technique developed by them proved to
be very useful when they were successful in
reproducing images of viscous fingering during a
dynamic displacement test8• They also reported
time derivatives of local composition and residual
oil distribution during waterflooding experiments.
Reservoir engineering research with CAT scanners
was further advanced by Cromwell et aI. 9 who
reproduced images of fluid distributions in Danianchalk and Berea sandstone.
. h' k d k 10-11More recently, WIt ~ac an co-wor ers
described techniques for rock property evaluation
and fluid flow visualization. They determined oilwater relative permeability by using CAT scanners
and the results were comparable to those determined
by conventional relative permeability measurement
techniques. Saturation results were within 2%whereas porosities were within 1% of those
determined by conventional techniques. Further
validation of saturation measurement techniques
with CAT scans was done by Manjanth12 who used the
CAT scan technology for verifying the Buckley
Leverett theory of immiscible fluid displacement.
Mohanty and Miller13 studied the factors affecting
unsteady-state relative permeabilities of a
mixed-wet reservoir rock using a CAT scanner. They
showed remarkable success in determining two-phase
relative permeability using this technology. Their
work revealed a number of regimes for the cross
sectional average saturation profile during
waterfloods at low rates. These regimes were
identified as frontal movement (up to 0.15 PV
injected), a non uniform saturation increase (up to
0.6 PV injected) and a uniform saturation increase.
Cross sectional averages of saturations determined
by CAT scanning revealed that the saturation
527
profiles determined by the JBN method at the outlet
of the core are incorrect during early. stages of
fluid displacement (up to 1 PV throughput).
Chatzis et al. 14 modified a medical CAT scanner to
study the saturation profiles during two- and
three-phase flow in gravity-assisted gas injection
processes. They reported relative success in
defining two-phase flow. The description of
three-phase flow was rather qualitative.
In order to investigate the effect of a third phase. I 1s d 'don capillary pressure, Dehgham et a. etermlDe
the height-saturation plot by CAT scanning using a
fourth generation GE 9800 scanner. They found
excellent agreement of the CAT scanning technique
with the conventional Hassler and Brunner method.
MacAllister et aI. 16 reported gas-oil relative
permeability data obtained through CAT scanning.
They found that two-phase relative permeability
could be very accurate since the in situ
saturation is being determined by solving twoequations only (one phase is measured directly, the
other by difference). However, the calculation of
three-phase relative permeability was found to be
questionable since one phase was estimated
independently. Consequently, determination of the
phase which was determined by difference was deemed
to be less accurate. However, in this study the
authors did not use the dual scanning approach.
Among others, they were able to observe non-uniform
saturation profiles during early stages of fluid
displacement. This was in agreement with Mohantyand Miller13 •
Hicks17 reported porosity and the distribution of
residual oil in heterogeneous carbonate cores from
the Fenn-Big Valley reservoir of Alberta and the
Taylor-Link Field of Texas. They also calculated
semivariographs and porosity distributions in order
to quantify rock heterogeneities.
Lenormand et aI. 18 verified a two-phase network
simulator by comparing numerical results with those
obtained with CAT scanning. Immiscible gas
injection behavior visualized by the CAT scanner
was in good agreement with numerical results.
6STATE-OF-THE-ART REVIEW OF NON-DESTRUCTIVE TESTING
WITH COMPUTER-ASSISTED TOMOGRAPHY SPE 2212;
Kantzas et al. 19 presented saturation profiles of
both imbibition and drainage immiscibledisplacements using a CAT scanner. They reported
difficulty in determining saturation profiles in a
three-phase flow system. Also, during two-phase
flow, they had to use dopants for better contrast
between fluids. This prohibited them from using
native fluids in the system. In one instance, the
residual oil saturation following a waterflood was
reported to be 20% as determined by CAT scan. Thiscompared to 38% as determined by material balance.Overall, their study showed the poorest agreement
so far reported on two-phase flow. Authors
provided no explanation for this poor measurement
with the CAT scanner.
Applications in Heavy Oil
Fransham and Jelen20 were the first researchers to
report flow visualization of a heavy oil
displacement using CAT scan technology. Their
technique allowed them to determine dynamic changes
in saturations during a waterflood of a heavy oil
saturated core. They also visualized fingeringphenomena both in macroscopic and microscopic
levels. Similar experiments were also reported bySedgwick and Miles-Dixon21
• More recently, Cuthielland Sedgwick? applied the CAT scanner to heavy oil
corefloods. They studied both steady-state and
dynamic waterlbitumen displacement tests using a
CAT scanner.
Reservoir Characterization
Honarpour et al.23 were one the first researchers
to use CAT scanners for reservoir characterization.
They determined rock heterogeneities, permeabilitybarriers, fractures and their orientations in North
Sea Danian chalk and Wyoming Phosphoria formations.
In the same line, Bergosh et al.24 used the CATscan technique for core analysis of naturally
fractured reservoirs. They found the techniquehelpful for characterizing fracture width, spacing,
tortuosity, interconnectivity and drilling mud
invasion. Honarpour et al. 25 further applied theCAT scan technology for reservoir characterization
studies. They used dolomite samples from the Upper
528
Madison Limestone Group. They successfully
identified mineral densities and their distributionwithin a fracture system. This application of CAT
for mineral identification was indeed novel. This
technique requires both density calibration and
contrast. They felt that in order to gain more
confidence, this technique should be compared with
other techniques. Unlike previous idea of rather
idealized fracture infilling, they observed that
fractures exhibited variable densities and
therefore multiple types of mineralization. Theydetected the presence of dolomite, calcite, quartz,
gypsum, anhydrite, illite, and chlorite minerals in
naturally occurring fractures. Bergosh and Lord26
conducted further studies on analysis of naturallyfractured reservoirs. They demonstrated that the
fracture porosity could be easily determined by CATscanning.
Kantzas et al. 19 recently presented a bulk of
information on seven Canadian reservoirs regarding
the use of reservoir characterization with a CAT
scanner. This was a follow-up study of a previous
investigation21• Their study could not be proven
to be useful for any quantitative characterization.
However, they provided useful information about theextent of core heterogeneities, presence offractures and fissures, etc. within the corelength.
A new method for imaging the three dimensionalmicrostructure of porous media was presented by
Jasti et aes. This method was based on high
resolution x-ray CT where a cone shaped divergingx-ray beam is used to generate two dimensional
transmission images. A three-dimensional
reconstruction array is created in this methoddirectly as opposed to series of two dimensional
slices as in conventional CAT. The authors
demonstrated the new technique using an
unconsolidated pack of glass beads. Even thoughthe technique appears to have great potential in
future, very little useful information was provided
at this stage of research.
SPE 22127 A. Bansal and M.R. Islam 7
Core Analysis and Well Logging
H 129unt et a . presented the ftrst detailed
technique for using CAT scan in core analyses.They noted that qualitative core analyses could bedone with relative ease using the CAT scan
technology. They showed application of CATscanning to
- determine the extent of drilling mud invasion
- detection of fractures
- characterization of preserved cores
screening of core prior to laboratory investi
gation
quantitative determination of porosity, permeability and mineralogy.
Kantzas et aeo reported modiftcation of the
hardware on a medical CAT scanner for adopting the
machine to the need of core analysis. They
modifted the CAT scanner for conducting
displacement tests on long cores and at any angle
of inclination. Using this machine, they studied
physical properties of porous rocks.
investigating damage to unconsolidated cores.
Auzerais et al.34 calculated densities and
effective atomic numbers for several anisotropic
rocks using CAT scanning as a part of a complete
core characterization project. Among others, theyconcluded that permeability variations are due to
variations in grain size and packing.
During an investigation into describing dual
porosity systems, Moss et al.35 presented an
algorithm for measurements of the constituent
porosities in a dual porosity matrix. They
saturated the core with a dopant and used an image
subtraction technique during scanning at differenttimes. Their technique is very similar to that of
Withjack et al. 36• Moss et al.35 suggested that
one can identify the easily accessible fractureporosity from the matrix porosity by following the
time sequence of the penetration process. As
examples, they applied the technique on coal and
shales. They observed that the width of the
porosity distribution is a good indicator of
heterogeneity in porosity. They also found that
porosity images can lead to statistically accurateporosity distributions.
One of the rare applications of CT scanning in
unconsolidated reservoir cores has been addressedby Kantzas et al.27
•32
• They used a medical x-ray
CAT scanner for characterizing a Canadian heavy oil
reservoir. Reservoir cores were collected and
rubber sleeves and were frozen in order to preserve
the form of the cores. Using a new post processing
software l, they determined a detailed image of
heterogeneities within each core section. These
heterogeneities included cracks, shale barriers,
and clay clasts. They also determined bulk densityas well as porosity of the reservoir cores. They
suggested one of the more interesting uses of CAT
scanner, Le. to identify locations in a core for
good plug selection. In addition, they reported
three- dimensional mapping of heterogeneities.
This line of work was further advanced by Gililand
and Coles33 who used the CAT scanner as a tool for
Narayanan and Deans3l used a
Deltascan 2020HR to validate
heterogeneous porous media.
fourth generation
a model for
529
Withjack et al.36 studied miscible displacement as
well as heterogeneities using a CAT scanner. The
focus of this study was to study heterogeneitycaused by permeability variation on a sub beddingplane scale. For the first time, they reported a
method for predicting core permeability using the
CAT scanner. The method is based on a bundle tube
model. In order to determine permeability, they
had to use results of miscible displacement tests.
This method assumes linear pressure drop along the
core. Such an assumption, even though a reasonableone for stratified heterogeneities, is not
satisfactory for vuggy and fractured cores. The
authors reported an excellent correlation betweenporosity and permeability.
Miscible Flooding and Mobility Control
CAT scanner was put into use for visualizing both
immiscible and miscible corefloods by Hove et
137-38 Th CAT . .a . e scannmg techmque allowed these
8
STATE-OF-THE-ART REVIEW OF NON-DESTRUCTIVE TESTINGWITH COMPUTER-ASSISTED TOMOGRAPHY SPE 22127
Cuthiel and Sedgwick22 studied two-phase
bitumen/water flow, steam flooding, and foam
propagation with the aid of a CAT scanner.
Withjack and Akervollll used the CAT scan
technology for studying miscible displacement in a
3-dimensional 5-spot scaled laboratory model. They
were very successful in describing miscible
displacement in a qualitative format.
presented suggestions for designing high-pressure
core materials for using CAT under realistic
reservoir conditions. Among others, they derived
fundamental equations for CAT scanning for
reservoir applications and listed examples of
applications of CAT in different areas, such as,
measurement of compressibility and compaction,
characterization of core materials, correlation of
core with well logs, multiphase core flood studies,
etc.
researchers to observe viscous fingering as well as
dispersion phenomena. They conducted the
corefloods using North Sea sandstones.
Some of the first applications of CAT in oil
research were done by the Shell Research Center of
Bellaire. Wellington and Vinegar4 used CAT scan
for studying mobility control with CO2
foam. They
used aluminum vessels for scanning high- pressure,
high-temperature core floods. It is well known
that aluminum exhibit very low x-ray attenuation.
Linear attenuation coefficients were used for
calculating in situ saturations. They used dual
energy scanning in order to image the three-phase
saturations. To obtain both density and effective
atomic number, CAT images are usually taken at two
x-ray energies. One energy is high enough for the
x-rays to be predominantly Compton-scattered and
one is low enough for them to be mostly
photoelectrically absorbed. Typically, a
combination of the two images for each pixel can be
used to generate separate images of bulk density
and effective atomic numbers. Wellington and
Vinegar4 used dual energy scanning to image the
three-phase fluid saturations in a core. Following
equations were used to measure saturations:
31 edNarayanan and Deans report
miscible displacement experiments in
carbonate rock samples under
conditions.
visualization of
highly vugular
high-pressure
miscible
A
cores.of the
They
with
Wellington and Vinegar39 studied mobility control
of CO injection using surfactant to generate foam.2
They reproduced images which identified several
mechanisms which explained the nature of mobility
control foam and how it renders CO2
stable against
gravitational and viscous forces during miscible
flood experiments.
Liu et al. 4O used the CAT scan technology to
monitor transient foam flow in laboratory linear
foam flood experiments. CAT scan results allowed
them to observe that the foam-displacement is not
piston-like. Gas channeling was observed near the
front whereas the foam eventually blocked thesechannels.
Hicks et al.41 studied vertical
displacements in heterogeneous carbonate
They were able to observe the propagation
displacement front using a vertical scanner.
compared numerical simulation results
experimental local and effluent concentrations.
Pu - P S + P S + /l S01 0 wI w gl g
/lu - P S + /l S + /l S02 0 w2w g2 g
S + S + S = 10 w g
where p is the total equivalent attenuation
coefficient of the core with several components
whereas Pw' Po' and P g are the attenuationcoefficients for the core saturation 100% with
water, oil, and gas, respectively. The subscripts
1 and 2 refer to energy level of 120 and 90 keV,
respectively. This was one of the first works
reported on detection of three phase flow with CAT
scanning under high temperature and pressure
conditions. In this study, 40% l-Iodododecane was
used as a dopant for the oil phase. In a later
study, Vinegar and Wellington3 outlined the theory
of CAT applications in reservoir engineering
studies. In this work, they discussed the optimum
energy settings and choice of dopants. They also
530
SPE 22127 A. Bansal and M.R. Islam 9
good agreement between these two results wasobtained.
ACKNOWLEDGEMENTS
REFERENCES
5. Kantzas, A., "A Theoretical and ExperimentalStudy of Gravity Assisted Inert Gas Injection asa Method of Oil Recovery", Ph.D. Thesis,University of Waterloo, Canada, 1988.
4. Wellington, S.L. and Vinegar, H.J., "CT Studiesof Surfactant-Induced CO
2Mobility Control",
paper SPE 14393 presented at the AnnualTechnical Conference an Exhibition of the SPE,Las Vegas, NV, 1985.
This study was made with partial financial support
of the Department of Petroleum Engineering,
University of Alaska, USA and the EMERTEC
Developments Inc., Calgary, Alberta, Canada.
S., and Cryte, C.C.,Tomography for theDisplacement in Porous
J., vol. 24 (1984) 53.
6. Wang, S.Y., Ayral,"Computer-AssistedObservation of OilMedia", Soc. Pet. Eng.
2. Waggener, R.G., Kereiakes, J.G., and Shalek,R.J., (eds.) "CRC Handbook of Medical PhysicsVolume IT", CRS Press (1984).
3. Vine¥ar, H.J. and Wellington, S.L., "TomographicImagmg of Three-Phase Flow Experiments", Rev.Sci. Instrum., vol. 58 (1), (1987) 96.
1. Kantzas, A., "Investigation of PhysicalProperties of Porous Rocks and Fluid FlowPhenomena in Porous Media Using ComputerAssisted Tomography", In Situ, vol. 14 (1)(1990) 77-132.
Withjack et al. 36 provided visualization of
miscible flooding in a well characterized Berea
core for a range of mobility ratios. Flow
visualization of the miscible displacement allowed
them to observe finger propagation in a
heterogeneous porous medium. They observed that
with increasing mobility ratios, displacements
become less dominated by local permeability
variations. A characteristic central finger was
observed for all high mobility ratios. They
concluded that displacement characteristics are lot
better understood by in situ saturation
determination through CAT scanning than that
provided by 2-D theory or by assumptions of
homogeneous systems.
Peters and Hardham42 conducted a series of
immiscible and miscible displacement tests in sand
packs for flow visualization. If a miscible
displacement is stable, they claimed that the
displacement data may be used to determine
longitudinal dispersion coefficient in a porous
medium. They also suggested that numerical models
could be calibrated by the saturation profiles
generated by a CAT scanner.
COMCLUSIONS
A historical review of the computer assisted
tomogrphy (CAT) as applied in the petroleum
industry is provided. Also, discussed is the
principle of CAT and basic density and saturation
measurement technques as applied to rocklfluid
systems. The review shows the usefulness of the
technique in a wide range of areas of petroleum
engineering and petroleum geology. CAT scan
technology continues to be improved and, as this
review reveals, a lot more progress has to be made
before a quantitatively useful information can be
derived in the areas of both reservoir
characterization and saturation profile monitoring.
However, even now the technique is unique in
testing samples in a much desired non-destructive
fashion.
7. Wang, S.Y., Ayral, S., Castellana, F.S., andCryte, C.C., "Reconstruction of Oil SaturationDistribution Histories During Immiscible LiquidLiquid Displacement by Computer-AssistedTomography", AIChE J, vol. 30(4) (1984) 642.
8. Wang, S.Y., Huang, Y.B., Pereira, V., and Cryte,C.C., "Application of Computer Tomography to OilRecovery from Porous Media" , Applied Optics,vol. 24 (23), (1985) 4021.
9. Cromwell, V., Kortum, D.J., and Bradley, D.J.,"The Use of a Medical Computer Tomography (CT)System to Observe Multiphase Flow in PorousMedia" , SPE 13098 presented at the AnnualTechnical Conference and Exhibition of the SPE,Houston, TJe, 1984.
10. Withjack, E.M., "Computed Tomography for RockProperty Determination and Fluid FlowVisualization", SPE 16951 presented at theAnnual Technical Conference and Exhibition ofthe SPE, Dallas, TJe, 1987.
531
10STATE-OF-THE-ART REVIEW OF NON-DESTRUCTIVE TESTING
WITH COMPUTER-ASSISTED TOMOGRAPHY SPE 22127
Bergosh, J.L., and Lord, G.D., "NewDevelopments in the Analysis of Cores fromNaturally Fractured Reservoirs", SPE 16805presented at the Annual Technical Conferenceand Exhibition of the SPE, Dallas, TX, 1987.
Kantzas, A., Chatzis, I., Macdonald, I.F., andDullien, F.A.L., "Using a Vertical Scanner forHorizontal Scanning in Non-Medical Applicationsof Computer Assisted Tomography" , CSNDTJournal, March-April (1988) 20.
Hunt, P.K., Engler, P., and Bajsarowicz, C.,"Computed Tomography as a Core Analysis Tool:Applications and Artifact Reduction Technique",SPE 16952 presented at the Annual TechnicalConference and Exhibition of the SPE, Dallas,TX, 1987.
Fransham, P.B. and Jelen, J., "Displacement ofHeavy Oil Visualization by CAT Scan" , CIM86-37-80 presented at the 37th Annual TechnicalMeeting of the Petroleum Society of CIM,Calgary, 1986; paper also published in the J.Can. Pet. Tech., vol. 26 (1987) 42-47.
"Newfrom
17479Reg.
Bergosh, J.L., and Lord, G.D.,Developments in the Analysis of CoresNaturally Fractured Reservoirs", SPE(abstract) presented at the SPE Calif.Meeting, Long Beach, CA, 1988.
Kantzas, A., Marentette, D.F., Erno, B., andHarding, S., "Characterization of a Heavy OilReservoir Using Computer Assisted Tomography ofCore Material" , paper presented at the ThirdPetroleum Conference of the South SaskatchewanSection of the Petroleum Society of CIM,Regina, Saskatchewan, Canada, 1989.
Jasti, J., Jesion, G., and Feldkamp, L.,"Microscopic Imaging of Porous Media UsingX-Ray Computer Tomography", SPE 20495 presentedat the Annual Technical Conference andExhibition of the SPE, New Orleans, LA, 1990.
Honarpour, M.M., McGee, K.R., Crocker, M.E.,Maerefat, N.L., and Sharma, B., "Detailed CoreDescription of a Dolomite Sample from the UpperMadison Limestone Group", SPE 15174 presentedat the Rocky Mountain Regional Meeting of theSPE, Billings, MT, 1986.
Sedgwick, G.E. and Miles-Dixon, E.,"Application of X-Ray Imaging Techniques to OilSands Experiments", J. Can. Pet. Tech., vol. 27(1988) 104-110.
Cuthiell, D. and Sedgwick, G., "X-ray CTApplied to Heavy Oil Corefloods", paper no. 9,presented at the Fourth UNITAR/UNDP Conferenceon Heavy Crude and Tar Sands, Edmonton, 1988.
Honarpour, M.M., Cromwell, V., Hatton, D., andSatchwell, R., "Reservoir Rock DescriptionsUsing Computer Tomography", SPE 14272 presentedat the Annual Technical Conference andExhibition of the SPE, 1985.
25.
29.
28.
27.
26.
24.
30.
23.
22.
21.
20.
14. Chatzis, I., Kantzas, A., and Dullien, F.A.L.,"On the Investigation of Gravity-Assisted InertGas Injection Using Micromodels, Long BereaCores and Computer-Assisted Tomography", SPE18284 presented at the Annual TechnicalConference and Exhibition of the SPE, Houston,TX, 1988.
13. Mohanty, K.K. and Miller, A.E., "FactorsInfluencing Unsteady Relative Permeability of aMixed-Wet Reservoir Rock", SPE 18292 presentedat the Annual Technical Conference andExhibition of the SPE, Houston, TX, 1988.
11. Withjack, E.M. and Akervoll, I., "ComputedTomography Studies of a 3-D MiscibleDisplacement Behavior in a Laboratory Five-SpotModel" , SPE 18096 presented at the AnnualTechnical Conference and Exhibition of the SPE,Houston, TX, 1988.
12. Manjnath, A., "Examination of the BuckleyLeverett Theory Using Computerized Tomography",SPE 17491 (abstract) presented at the SPECalifornia Reg. Meet., Long Beach, CA, 1988.
15. Dehghani, K., Bansal, A., Ogbe, D.O., andOstermann, R.D., "The Effect of the Presence ofa Third Phase on Capillary Pressure byCentrifuge Method and CT Scanning", SPE 18793presented at the SPE California Reg. Meet.,Bakersfield, CA, 1989.
16. MacAllister, D.J., Miller, K.C., Graham, S.K.,and Yang, C.T., "Application of X-Ray CTScanning to the Determination of Gas-WaterRelative Permeabilities" , SPE 20494 presentedat the Annual Technical Conference andExhibition of the SPE, New Orleans, LA, 1990.
17. Hicks, P.J., Jr., Deans, H.A., and Narayanan,K., "Experimental Measurement of theDistribution of Residual Oil Saturations inHeterogeneous Carbonate Cores Using X-RayComputerized Tomography" , CIM/SPE 90-68presented at the International TechnicalMeeting of the CIM and SPE, Calgary, Alberta,Canada, 1990.
18. Lenormand, R., Kalaydijan, F., Bieber, M.T.,and Lombard, J.M., "Use of MultifractalApproach for Multiphase Flow in HeterogeneousPorous Media: Comparison with CT-ScanningExperiment", SPE 20475 presented at the AnnualTechnical Conference and Exhibition of the SPE,New Orleans, LA, 1990.
19. Kantzas, A., Marentette, D.F., and Jha, K.N.,"Computer Assisted Tomography: FromQuantitative Visualization to Quantitative CoreAnalysis", paper CIMIAOSTRA 91-72 presented atthe Technical Conference of the CIMIAOSTRA,Banff, Alberta, Canada, 1991.
532
SPE 22127 A. Bansal and M.R. Islam 11
31. Narayanan, K. and Deans, H.A., "A Flow ModelBased on the Structure of Heterogeneous PorousMedia", SPE 18328 presented at the AnnualTechnical Conference and Exhibition of the SPE,Houston, TX, 1988.
32. Kantzas, A., Marentette, D.F., and Skalinski,M., "Computer Assisted Tomo~raphy as aComplementary Tool to Well Loggmg", paper P,presented at the 12th Formation EvaluationSymposium of the Canadian Well Logging Society,Calgary, Alberta, Canada, 1989.
33. Gililand, R.E. and Coles, M.E., "Use of CTScanning in the Investigation of Damage toUnconsolidated Cores", SPE 19408, presented atthe SPE Formation Damage Control Symposium,Lafayette, LA, 1990.
34. Auzerais, F.M., Ellis, D.V., Luthi, S.M.,Dussan, E.B.V., and Pinoteau, B.J., "LaboratoryCharacterization of Anisotropic Rocks" , SPE20602 presented at the Annual TechnicalConference and Exhibition of the SPE, NewOrleans, LA, 1990.
35. Moss, R.M., Pepin, G.P., and Davis, L.A.,"Direct Measurement of the ConstituentPorosities in a Dual Porosity Matrix" , paper9003, presented at the Fourth Annual TechnicalConference of the Society of Core Analysis,Dallas, TX, 1990.
36. Withjack, E.M., Graham, S.K., and Yang, C.T.,"Determination of Heterogeneities and MiscibleDisplacement Characteristics in Corefloods byCT Scanning", SPE 20490 presented at the AnnualTechnical Conference and Exhibition of the SPE,New Orleans, LA, 1990.
37. Hove, A.O., Ringen, J.K., and Read, P.A.,"Visualization of Laboratory Corefloods withthe Aid of Computerized Tomography of X-rays",SPEFE, vol. 2 (1987) 148; paper originallypresented as SPE 13654 at the SPE CaliforniaReg. Meet., Bakersfield, CA, 1985.
533
38. Hove, A.O., Nilsen, V., and Jenkes, A.,"Visualization of Xanthan Flood Behavior inCore Samples by Means of X-ray Tomography", SPE17342 presented at the SPE/DOE Enhanced OilRecovery Symposium, Tulsa, OK, 1988.
39. Wellington, S.L. and Vinegar, H.J.,"Surfactant-Induced Mobility Control for CarbonDioxide Studied with Computerized Tomography",Surfactant-Based Mobility Control, Chapter 17,ACS, 1988.
40. Liu, D., Castanier, L.M., and Brigham, W.E.,"Analysis of Transient Foam Flow in I-D PorousMedia with CT" , SPE 20071 presented at theSPE California Regional Meeting, Ventura, CA,1990.
41. Hicks, P.J., Jr., Narayanan, R., and Deans,H.A., "An Experimental Study of MiscibleDisplacements in Heterogeneous Carbonate CoresUsing X-Ray CT", SPE 20492 presented at theAnnual Technical Conference and Exhibition ofthe SPE, New Orleans, LA, 1990.
42. Peters, E.J. and Hardham, W.D., "Visualizationof Fluid Displacements in Porous Media UsingComputed Tomography Imaging", J. Pet. Eng.Sci., vol. 4 (1990) 155.