An Advanced Special Core Analysis Study for a Middle Eastern Carbonate Field (SPE 161932)

SPE 161932
An Advanced Special Core Analysis Study for a Middle Eastern Carbonate
Field

Lisa Lun, Kyle Guice, Jim Kralik, ExxonMobil Upstream Research Co.; Jon Meissner, ExxonMobil Production Co.;
Maher Kenawy, Zubair Kalam, and Taha al-Dayyni, Abu Dhabi Co. for Onshore Oil Operations
Copyright 2012, Society of Petroleum Engineers
This paper was prepared for presentation at the Abu Dhabi International Petroleum Exhibition & Conference held in Abu Dhabi, UAE, 11–14 November 2012.
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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
A novel special core analysis (SCAL) study was conducted utilizing samples from a Middle Eastern Carbonate Reservoir in
order to gain insights into flow behavior across stylolitic intervals. This study included relative permeability and capillary
pressure measurements performed on individual core plug and core plug composite samples, as well as a unique waterflood
experiment on a four-inch diameter whole core composite. All laboratory flow measurements were performed at reservoir
conditions of temperature, pressure, and net confining stress. As part of this study, it was demonstrated that wettability
restoration remains a significant challenge for carbonate core samples, with implications for coring and core analysis
program design and interpretation of historic SCAL data. Core-scale simulation using measured relative permeability and
capillary pressure data along with whole core rock properties provides an opportunity to validate laboratory results across
laboratory scales and can also serve as an intermediate to mechanistic modeling studies at larger scales. In this paper, the
novel technical approach and significant findings for the special core analysis study are presented, with implications for
modeling of displacement processes across stylolitic intervals in complex carbonate reservoirs. General recommendations for
the design of special core analysis programs are also presented.
Introduction
High quality special core analysis (SCAL) data (e.g., relative permeability and capillary pressure), integrated into internallyconsistent saturation functions, provide critical inputs into reservoir performance prediction throughout the reservoir lifecycle. Acquiring high-quality data in the laboratory generally requires that measurements must be on rock samples
representative of the reservoir (the right samples); measurements must be made under conditions representative of
displacement processes in the reservoir (the right conditions); measurements must be made using precision equipment and
techniques (the right equipment); and trained and experienced technologists are needed to conduct the measurements and

model the data (the right people).
For certain carbonate rock types, ensuring that measurements are performed on rock samples representative of the reservoir
presents a challenge of scale. A core sample may be considered representative of a specific reservoir rock type if it contains
the relevant features that distinguish the reservoir rock type in a distribution that is “homogeneously heterogeneous” at the
selected measurement scale. Increasing the size of the measured core sample is often a good way to achieve a representative
sample if key features are not homogeneously distributed at smaller scales. For some carbonate rock types, including highly
karsted rock types, identifying core samples with a representative distribution of heterogeneities at any laboratory scale is a
significant challenge.
In these cases, one can gain insights into the impact of select heterogeneities on special core analysis measurements by
performing an integrated suite of measurements spanning multiple length scales. Given that the objective of any special core
analysis study is to deliver high quality data to address questions at scales larger than can be measured in the laboratory, the
desired multi-scale approach is to supplement special core analysis measurements at traditional scales with additional
measurements at larger length scales. As part of an integrated suite of measurements, whole core testing is particularly useful


2

SPE 161932

to address the effects of key features of interest on flow performance, as a whole core represents the largest sample size that
can be acquired from a reservoir.
In this study, an integrated suite of special core analysis measurements spanning multiple length scales is employed to
investigate flow behavior across stylolitic features for a Middle Eastern carbonate reservoir. Stylolites are diagenetic features
that occur in many carbonate reservoirs as a result of pressure dissolution, generally result in low permeability and low
porosity due to cementation in the immediate vicinity of a stylolite seam, and exist with significant variability in
permeability, porosity, spatial extent, and connectivity. At reservoir scales, stylolitic features may behave as either barriers or
baffles to vertical flow. A better understanding of flow behavior across stylolitic intervals is desired to better interpret
production data and support future well planning.
The study presented was divided into three parts. First, special core analysis measurements were performed on representative
core samples at representative reservoir conditions using a multi-scale approach. Laboratory measurements included capillary

pressure and relative permeability measurements performed on core plug and core plug composite samples, as well as a
unique whole core study to quantify flow across a large stylolite interval in the laboratory. Second, the multi-scale laboratory
measurements were integrated into a core-scale simulation model of the whole core experiment. Finally, validated special
core analysis inputs from the core-scale simulation model were used as part of a mechanistic model to understand flow across
a stylolitic interval at reservoir scales.
Measurement of Capillary Pressure and Relative Permeability on Core Plugs and Core Plug Composites
Capillary pressure and steady-state relative permeability measurements were performed on core samples in their native state
(preserved) at reservoir net confining stress. Samples were selected to cover a range of reservoir rock types and rock
properties. The testing strategy included relative permeability and capillary pressure measurements on stylolitic plug samples.
The absolute permeability of the plugs used in this study ranged from 1 mD to 60 mD.
Water-oil imbibition and secondary drainage capillary pressure measurements were performed by the centrifuge method on
individual, 1.5-inch diameter plug samples using rotovaped stock tank oil and synthetic reservoir brine at a temperature of
170°F. Primary drainage capillary pressure measurements were performed by the centrifuge method with nitrogen gas
(displacing phase) and synthetic brine (displaced phase). Relative permeability measurements were performed by the steadystate method on two-inch diameter samples (either long vertical plugs or core plug composites) using live reservoir fluids at
reservoir temperature and pressure.
All capillary pressure and relative permeability measurements presented in this paper are plotted against normalized water
saturations in a similar manner as presented in the Appendix of Meissner et al. (2009). For primary drainage capillary
pressure, the normalized water saturation, Swn, is defined as:

S wn =

S w ! S wirr
1 ! S wirr

where Sw is the water saturation and Swirr is the irreducible water saturation of each sample. For primary imbibition and
secondary drainage capillary pressure, Swn is defined as:

Swn =

Sw ! Sw,min

1 ! Sw,min ! Sorw

where Sw,min is the minimum water saturation achieved in that cycle for each sample and Sorw is residual oil saturation of that
sample. All of the relative permeability curves were scaled in a similar manner as the primary imbibition capillary pressure
curve except Sw,min is the minimum water saturation achieved over all hysteresis cycles for each sample and Sorw is the
minimum oil saturation achieved over all cycles for each sample.
Results from centrifuge capillary pressure measurements are presented in Figures 1-3. Water-oil primary imbibition capillary
pressure data are plotted in Figure 1. Oil-water secondary drainage capillary pressure data are plotted in Figure 2. Gas-water
primary drainage data are plotted in Figure 3. In all cases, the capillary pressure data are calculated at the inlet face of the
core plug using a numerical technique to solve the general equation described by Hassler and Brunner (1945). The same
samples set of seven samples were used for all capillary measurements, although all cycles were not performed on Sample 3
(secondary drainage omitted) or Sample 7 (primary imbibition and secondary drainage omitted).
Water-oil relative permeability measurements were conducted using the steady-state method on two-inch diameter samples
(either long vertical core plugs or core plug composites) at full reservoir conditions (i.e., temperature, pressure, net confining
stress). The steady-state method allows for accurate measurement of oil and water relative permeabilities over a wide range
of fluid saturations. The fully-recirculating experimental apparatus used for steady-state relative permeability measurements


SPE 161932

3

was previously described in Braun and Blackwell (1981). Measurements were performed on each sample in the primary
imbibition, secondary drainage, and secondary imbibition saturation pathways. For each saturation pathway, the waterflood
(imbibition) or oilflood (drainage) data acquired after the final steady-state fractional flow were analyzed as an unsteady-state
flood utilizing the method of Johnson, Bossler, and Naumann (1959).
Water-oil relative permeability data for three samples are presented on normalized saturation scales in Figures 4-6. In each
figure, steady-state measurements are presented as filled circles (!, krow) and filled triangles (!, krw), and unsteady-state
flood data are presented as solid lines. Each figure includes primary imbibition, secondary drainage, and secondary
imbibition pathways. Hysteresis in both oil and water relative permeability data are generally observed between primary

imbibition and secondary drainage for these samples, and is slightly more pronounced for the water phase. Hysteresis is a
qualitative indicator of wettability, and is usually most pronounced for the non-wetting phase. The hysteresis data suggest
that all samples are mixed-wet with a slight wettability preference for oil.
Following the relative permeability and capillary measurements described above, the samples were cleaned and restored, and
capillary pressure and relative permeability measurements were again performed on the same samples. Wang (1994) and
Meissner et al. (2009) have observed in independent studies that wettability restoration methods for carbonate samples are
not reliable, particularly for lower quality samples (i.e., below 100 mD). A description of the cleaning process is included in
the Appendix.
Representative examples of restored-state results of capillary pressure (primary imbibition and secondary drainage) and
relative permeability (primary drainage) are shown in Figures 7-9 for Samples 5, 2, and 8 respectively. The data sets have
been normalized on the same basis for each sample. We observed marked differences between the results for preserved and
restored-state samples in both capillary pressure and relative permeability with the exception of Sample 2 (Figure 8). For this
sample, the capillary pressure results appear more reproducible. The absolute permeability of Sample 2 (~60 mD) was the
highest in the sample set. Meissner et al. (2009) presented data that indicated that wettability restoration for carbonate rocks
may be more reliable on higher quality samples, and the results of this study are consistent with this observation.
While wettability itself is not a direct input into most reservoir simulators, high-quality relative permeability and capillary
pressure data require that measurements are performed on samples in a condition representative of the reservoir. Accordingly,
since carbonate wettability restoration remains a significant challenge, careful coring and core preservation efforts are
generally required to produce high-quality special core analysis data for carbonate reservoirs.
Whole Core Waterflood Experiment
In addition to relative permeability and capillary pressure measurements conducted on core plug and composite samples, a
novel whole core experiment was designed to study flow across a stylolitic whole core segment. The experiment was
performed on a 21.6-inch long, four-inch diameter whole core composite at reservoir temperature, pressure, and net confining
stress in a unique whole core apparatus. The whole core composite consisted of four whole core segments in their preserved
state. Each segment was selected to represent an appropriate reservoir interval. A schematic of the whole core composite is
shown in Figure 10. The second segment from the top, Sample WC2, contained a visible stylolitic interval. Pressure taps
were located at several points along the composite, as illustrated in Figure 10.
A waterflood was performed vertically downward, and oil production and the pressure gradients were monitored. In addition,
in-situ saturation monitoring was utilized to measure water saturation along the length of the whole core composite
throughout the waterflood. The waterflood was performed at three progressively increasing flow rates (0.2, 0.4, 0.8 cc/min).

A total of 21.6 pore volumes of brine were injected over the course of the experiment.
Oil production and pressure drop data from the whole core waterflood experiment are presented in Figure 11. Oil production
is represented as a dark blue line, and total pressure drop is represented as a red line. Local pressure drop data (i.e., between
individual pressure taps) are also shown in orange (P1-P2), blue (P2-P3), and green (P3-P4). The water saturation profile along
the length of the whole core composite is presented in Figure 12 for various points during the waterflood.
The oil production data indicate water breakthrough after 620 cc of oil production. The movement of the flood front before
breakthrough can be clearly observed from the saturation profile data (Figure 12) and also from the progression of peaks in
the local pressure drop data with increasing distance from the inlet of the core (Figure 11). The center section exhibits the
largest pressure drop as the flood front passes (27 psi), followed by the bottom section (17 psi) and the top section (7.4 psi).
While the top section is approximately half the length of the other two sections, the center and bottom sections are of similar
size. The difference in peak pressure drop can be attributed to the presence of the lower permeability stylolitic interval in the
center section. After breakthrough, there is a period ("Qi = 400 cc) in which no additional oil is produced from the whole
core composite. This behavior after breakthrough is also reflected in the in-situ saturation monitoring data and is likely the


4

SPE 161932

result of baffling within the stylolitic interval. Subsequently, oil production recommences and then becomes progressively
slower throughout the waterflood.
The water injection rate was doubled twice to see if there was a rate dependency on oil production. First, the rate of water
injection into the core was doubled after 12.4 pore volumes of brine had been injected into the whole core composite. Then
the rate of water injection was doubled again after another 16.2 total pore volumes were injected. In each instance, there is an
increase in the total and localized pressure drops, as shown in the figure. Oil production from the whole core composite
appears to be independent of injection rate, provided that total production is evaluated relative to throughput (rather than
time). After the second increase in the injection rate, there is a notable decline in the pressure drop across the bottom section
(P3-P4), whereas the local pressure drops across the other sections remain constant. A decrease in pressure drop across the
bottom of the core at an elevated injection rate is suggestive of a capillary end effect at the bottom of the core.
Water saturation profiles are shown at several points during the waterflood in Figure 12. The stylolitic whole core segment

reaches a final water saturation that is notably higher (>95%) than the remainder of the whole core composite. The location
of the flood front is clearly visible for all saturation profiles before water breakthrough. The whole core waterflood
experiment clearly demonstrated that the selected stylolitic interval does not present a barrier to vertical flow at the whole
core scale, that oil can be effectively displaced across the stylolite interval and that low residual oil saturations can be
ultimately obtained above, below, and within the stylolite interval.
Core-scale simulation
A core-scale simulation model was constructed to reproduce the results of the whole core waterflood experiment and verify
the saturation function inputs measured on plugs and plug composites. The whole core waterflood experiment was modeled
using EMpower, the ExxonMobil proprietary reservoir simulation software (Beckner et al. 2001). Saturation functions
developed from the capillary pressure and relative permeability measurements on plugs/plug composites were used as inputs
to the model. The goal of the simulation was to validate the saturation function inputs from the plug scale experiments
(capillary pressure and relative permeability) for the different rock types represented in the whole core experiment.
Cumulative oil production and pressure drop across the core were matched, although a rigorous history match was not the
goal of the study. Instead matching the overall behavior was deemed sufficient.
The whole core waterflood is modeled using two fluid phases (black oil model) with a rectangular model constructed of the
following total dimensions: 81 cm2 (equivalent to 4 inch diameter cross-sectional area), with a 10 # 10 gridding and 112
layers in the vertical dimension, with each vertical layer having a total height of 0.5 cm. The vertical resolution of the grid
was based on the resolution of the in-situ saturation monitoring system used in the whole core experiment (0.5 cm). The
reference depth for the uppermost vertical layer is 8000 ft. Each layer has a constant porosity and permeability based on
porosity and permeability maps that were independently acquired on the whole core composite. Absolute permeability was
rescaled to permeability of oil at connate water, kocw, to match the reference permeability used to scale the relative
permeability curves. Horizontal permeability was set equal to vertical permeability based on whole core measurements of
horizontal and vertical permeability. Displacement data (e.g., capillary pressure and relative permeability) were applied
separately for each whole core segment based on the results of plug and core plug composite measurements. An injector well
was placed in the uppermost layer to facilitate water injection in the model, and a producer well was connected to the
lowermost layer in the model. The porosity and absolute permeability of the injector and producer layers were matched to
adjacent rock layers to improve numerical stability. Capillary pressure in the producer layer was set to zero to be consistent
with capillary effects observed at the outlet of the whole core composite during the waterflood. Injection rate limits and
producer pressure limits based on experimental data were used as boundary limits in the simulation model. Layers 1, 14, 64,
and 112 (arranged from top to bottom) correspond to the locations of the pressure transducers P1, P2, P3, and P4 respectively.

The initial water saturation of the whole core composite was slightly lower than that of the plug samples used in the capillary
pressure and steady-state relative permeability measurements. A design of experiments approach was used to investigate
various approaches for endpoint scaling of capillary pressure and relative permeability functions in the core-scale simulation
model, and also to assess the model’s sensitivity to the shape of the positive branch of the primary imbibition capillary
pressure curves, which was not directly measured in the laboratory. The various simulation runs revealed that the simulation
results were not sensitive to endpoint scaling of the capillary pressure curves or the positive branch of the primary imbibition
capillary pressure curves. For relative permeability endpoint scaling, the optimal method to match the whole core waterflood
experiment was to extend the Corey curves fit to laboratory data to the lower whole core Swi, with appropriate rescaling in
krow and krw such that the initial oil permeability (kocw) in the whole core simulation matched that of the experiment.
Inclusion of secondary drainage relative permeability and capillary pressure data was considered in the model, but a
sensitivity to capillary pressure hysteresis revealed minimal impact on results (i.e., less than 1% difference in pressure or
cumulative oil production match). Accordingly, hysteresis was not included in the final core-scale simulation model.


SPE 161932

5

A comparison of results, oil production and pressure drop, from the whole core waterflood and the core-scale simulation
model are presented in Figure 13 where experimental results are in solid lines and simulation results are in dashed lines.
Results are plotted with respect to volume of brine injected. Overall, we find reasonable agreement between the experiment
and core scale simulation results. There are noticeable differences in pressure drop in the lower portion of the composite (P3P4) and water breakthough. During the first two flow periods, P3-P4 is lower in the simulation than in the experiment while in
the third flow period P3-P4 is higher in the simulation. Also, water breakthrough is predicted as occurring later in the
simulation than what occurred in the experiment. This delay may be related to more complex heterogeneity existing in the
stylolite, such as baffles, and in the rest of the core than were effectively captured in this one-dimensional model.
Screenshots from the simulation of water saturation in the core composite are shown in Figure 14 (green is low water
saturation and blue is high water saturation) and line plots in Figure 15. The water front proceeds uniformly through the core,
which is expected given the simplicity of the simulation model. Water saturations at the front are higher by approximately 10
saturations units compared to the experimental data. Water saturations at the end of the flood match the experimental data
better (approximately 75-80% in the Upper/Lower Zones and 85-90% in the stylolitic interval). The simulation showed a

small amount of capillary end effect at the bottom, outlet of the composite agreeing with the experimental data. The water
saturation in the stylolitic interval was higher than in other parts of the core, which is consistent with the experimental results.
While there are some differences between simulation and experimental results, the match is of acceptable quality given the
purpose of simulation and the simplicity of the model. The goal of the core scale simulation study was not to achieve a
rigorous history match. Instead, the goal was to obtain a reasonable description of the physical behavior to validate the
saturation function inputs and this was achieved using a fit-for-purpose simulation model.
Mechanistic model
The final phase of the project involved taking validated inputs to the core scale model and applying them at a mechanistic
model scale. The mechanistic model was built in EMpower. The grid had a 4 km # 1 km areal span and was 0.055 km thick
(180.5 ft). The model was built as a “layer-cake” with constant porosity and permeability values in each layer. The layers
were grouped into three gross divisions: an upper zone, a stylolitic interval, and a lower zone. The stylolitic interval was
explicitly modeled with horizontal permeability on the order of 0.5 mD. The horizontal permeability of the lower zone was
on the order of 10 mD while the permeability of the upper zone was approximately 50-100 times greater. The vertical
permeability was set equal to the horizontal permeability based on routine permeability measurements done on whole core
samples. Relative permeability and capillary pressure curves used in the mechanistic model are the same as those presented in
the core-scale model of the whole core experiment. There were three horizontal wells located at the bottom of the lower zone:
one injector and two producers. The producers were spaced from the injector at 2 km and 4 km. Water was injected at a rate
of 10,000 bbl/day with a constraint that the injector bottom hole pressure not exceed 6500 psi. Producers were maintained at
constant pressure boundary conditions of 1500 psi for producer1 and 1200 psi for producer2 which were above the bubble
point pressure of 1000 psi. Each producer was also set to shut-in when the produced water cut exceeded 80%.
Figure 16 shows screen captures of the oil saturation in the mechanistic model at various times where blue is low oil
saturation and green is high oil saturation. Despite injector and producer placement in the lower zone, the upper zone is swept
faster than the lower zone in the mechanistic model. The preferential sweep in the upper zone is primarily due to the large
contrast in horizontal permeability between the upper and lower zones. Gravitational cross flow from the upper to lower
zones (light blue color below the stylolite) indicates that the stylolitic interval acts as a weak baffle to flow in the vertical
direction, which is supported by observations in the laboratory study. The mechanistic model is a useful tool that can be used
to quickly assess the impact of different parameters (well placement, permeability contrast, baffles, saturation functions, etc.)
on a reservoir scale.
Conclusions
A novel workflow has been presented to measure and validate high-quality special core analysis data for heterogeneous

carbonates. The workflow consists of the following components:
• An integrated suite of high-quality special core analysis measurements (relative permeability and capillary pressure)
performed on core plug and core plug composite scales.
• A whole core experiment designed to assess the impact of specific heterogeneities that are not homogeneously
distributed in a core sample at the laboratory scale.
• Core-scale simulation of the whole core experiment to validate special core analysis results at a larger scale.
• Implementation of validated special core analysis results into mechanistic or field scale models.
Achieving high-quality special core analysis data generally requires that measurements must be performed on representative
rock samples, at representative reservoir conditions, using precision equipment and techniques, and are planned, executed,
and integrated by trained and experienced technologists. For carbonate core samples, there are particular considerations to


6

SPE 161932

note:
•
•

Wettability restoration to achieve representative reservoir conditions has been demonstrated in this and other studies
to remain a significant challenge. Accordingly, careful planning of carbonate coring and core preservation steps is
required to ensure native state rock samples are available for high quality SCAL measurements.
For highly heterogeneous carbonates, identifying representative rock samples should include consideration for scale.
In this study, a multi-scale approach is demonstrated that includes flow studies at a maximum laboratory scale (i.e.,
whole core).

This workflow was successfully demonstrated for a stylolite-containing Middle Eastern carbonate reservoir. The study
revealed that, at laboratory scales, stylolitic features do not always act as barriers to the displacement of oil by water but
rather act as flow baffles. High-quality special core analysis measurements acquired on core plug and core plug composite

samples have been validated against a core-scale simulation of a whole core waterflood experiment and were used directly in
a mechanistic model to describe flow across a stylolitic interval at a field scale.
Acknowledgments
The authors wish to thank Abu Dhabi Co. for Onshore Oil Operations and ExxonMobil Upstream Research Company for
their support and permission to publish this paper. The authors also would like to recognize M.M. Honarpour, Abi Modavi,
J.A. Boros, C. Chiasson, D.C. Laverick, R. Longoria, L.J. Manak, L.J. Poore, J. Rainey, and A.C. Wood for their significant
contributions.
Nomenclature
cc
milli-Liters
cc/min milli-Liters per minute
cm
centimeters
square centimeters
cm2
EMpower proprietary ExxonMobil reservoir simulator
ft
feet (length)
km
kilometers
permeability to oil at connate water
kocw
relative permeability to oil (displaced by water)
krow
relative permeability to water
krw
mD
milli-Darcies
pressure tap, where n = 1-4
Pn

psi
pounds per square inch
!Qi
change in volume injected
SCAL special core analysis
Sorw
residual oil saturation
water saturation
Sw
Swi
initial water saturation
irreducible water saturation
Swirr
Sw,min minimum water saturation
normalized water saturation
Swn
USBM United States Bureau of Mines
References
Beckner, B.L., Hutfilz, J.M., Ray, M.B., and Tomich, J.F. 2001. EMpower: ExxonMobil’s New Reservoir Simulation System.
Paper SPE 68116 presented at the 2001 SPE Middle East Oil Show, Bahrain, 17-20 March 2001.
/>Braun, E.M., and Blackwell, R.J. 1981. A Steady-State Technique for Measuring Oil-Water Relative permeability Curves at
Reservoir Conditions. Paper SPE 10155 presented at the 56th Annual Fall Technical Conference and Exhibition, SPE of
AIME, San Antonio, Texas, 5-7 October 1981. />Corey, A.T. and Rathjens, C.H. 1956. Effect of Stratification on Relative Permeability. J Pet Technol 8 (12): 69-71. SPE744-G. />Gomes, J.S., Ribeiro, M.T., Strohmenger, C.J., Negahban, S., Kalam, M.Z. 2008. Carbonate Reservoir Rock Typing – The
Link between Geology and SCAL. Paper SPE 118284 presented at the Abu Dhabi International Petroleum Exhibition and
Conference, Abu Dhabi, UAE, 3-6 November, 2008. />

SPE 161932

7

Hassler, G.L. and Brunner, E.M. 1945. Measurement of Capillary Pressure in Small Core Samples. Petroleum Transactions,
AIME 160: 114-123. />Honarpour, M.M., Nagarajan, N.R. and Sampath, K. 2006. Rock/Fluid Characterization and Integration - Implications on
Reservoir Management. Distinguished Author Series, J Pet Technol 58 (9): 120-130. SPE-103358-MS.
/>Honarpour, M.M., Djabbarah, N.F., and Sampath, K. 2005. Whole-Core Analysis – Experience and Challenges. SPE Res
Eval & Eng 8 (6): 460-469. SPE-81575-PA. />Honarpour, M.M., Djabbarah, N.F., Kralik, J.G. 2004. Expert-based Methodology for Primary Drainage Capillary Pressure
Measurements and Modeling. Paper SPE 88709 presented at the 11th Abu Dhabi International Petroleum Exhibition and
Conference, Abu Dhabi, UAE, 10-13 October, 2004. />Ingle, J.D. and Crouch, S.R. 1988. Spectrochemical Analysis. New Jersey: Prentice Hall.
Johnson, E.G., Bossler, E.M., and Naumann, V.O. 1959. Calculation of Relative Permeability for Displacement Experiments.
Petroleum Transactions, AIME 216: 370-372.
Kalam, M.Z., Al-Alawi, S.M., Al-Shekaili, S. 1997. A Novel Technique for Predicting End-Point Relative Permeabilities of
Heterogeneous Limestones from Log Derived Input Data. Paper SPE 37694 was presented at the 1997 Middle East Oil Show
and Technical Exhibition, Bahrain, 15-18 March, 1997. />Koepnick, R.B. 1987. Distribution and Permeability of Stylolite-Bearing Horizons Within a Lower Cretaceous Carbonate
Reservoir in the Middle East. Paper SPE 14173 presented at the SPE Annual Technical Conference and Exhibition, Las
Vegas, NV, 22-25 September, 1985. />Kralik, J.G., Guice, K., and Meissner, J.P. 2010. Methods and Tools for the Development of Consistent Reservoir Rock Type
Based Relative Permeability and Capillary Pressure Models for Reservoir Simulation. Paper SPE 137357 presented at the
Abu Dhabi International Petroleum Exhibition & Conference (ADIPEC), Abu Dhabi, UAE, 1-4 November 2010.
/>Meissner, J.P., Wang, F.H.L., Kralik, J.G., Ab Majid, M.N., Bin Omar, M.I., Attia, T., Al-Ansari, K. 2009. State of the Art
Special Core Analysis Program Design and Results for Effective Reservoir Management, Dukhan Field, Qatar. Paper IPTC
13664 presented at the International Petroleum Technology Conference (IPTC), Doha, Qatar, 7-9 December 2009.
/>Montgomery, D.C, Design and Analysis of Experiments, 7th Edition, New York: John Wiley & Sons, 2008.
Serag El Din, S. Dernaika, M.R., Kalam, M.Z. 2011. The Impact of Heterogeneity and Multi-Scale Measurements on
Reservoir Characterization and STOOIP Estimations. SPE Paper 147950 presented at the SPE Reservoir Characterization
and Simulation Conference and Exhibition, Abu Dhabi, UAE, 9-11 October 2011. />Wang, F.H. 1994. Some Aspects of Wettability Alteration, Restoration, and Preservation. In Proceedings of the 3rd
International Symposium on Evaluation of Reservoir Wettability and its Effect on Oil Recovery, 21-23 September 1994,
Laramie, WY, ed. N.R. Morrow, 119-122.
Wang, F.H. 1998. Effect of Wettability Alteration on Water/Oil Relative Permeability, Dispersion, and Flowable Saturation
in Porous Media. SPE Res Eng 3 (2): 617-628. SPE-15019-PA. />Appendix – Cleaning Procedure
After capillary pressure and relative permeability measurements were run on the plug samples, the plugs were extracted,
cleaned, and then wettability was restored. A different solvent mixture and sequence was used to clean the plugs compared to
the method reported in Wang (1988) and Meissner et al. (2009), however with comparable temperatures and flow rates. In

this study, the solvent sequence used was toluene saturated with water, then a methanol/acetone/chloroform azeotrope, and
finally followed by methanol only. The remaining methanol was blown out and dried under active vaccum.


8

SPE 161932

Figures
#

#(&

#("

#()

#(*

&$

'

#

&#
!%

&"
+,-./01'

!"#$%%"&'()&*++,&*-(#+$

!"#$%%"&'()&*++,&*-(#+$

!'#

+,-./01&

!'%

+,-./01$
+,-./01"

!&#

+,-./01%

+,-./01)

!&%

&!

()*+,-.&
()*+,-."

%

()*+,-.#

()*+,-./

$

()*+,-.$

#
!$#

"
!$%

!
!

!"#

./&0"%$1*2(3"4*&(5"4,&"4$/6-(7&"84$/6

Figure 1. Water-oil primary imbibition capillary pressure
data.
&$!
&#!
()*+,-.&

&"!

!"#$%%"&'()&*++,&*-(#+$

()*+,-."

()*+,-./

&!!

()*+,-.#
%!

()*+,-.0
()*+,-.$

$!

()*+,-.1

#!
"!
!
!

!'"

!'#

!'$

!'%

!'"

!'#

!'$

!'%

&

./&0"%$1*2(3"4*&(5"4,&"4$/6-(7&"84$/6

&

./&0"%$1*2(3"4*&(5"4,&"4$/6-(7&"84$/6

Figure 3. Gas-water primary drainage capillary pressure
data.

Figure 2. Oil-water secondary drainage capillary pressure
data.


SPE 161932

9

'"#

!"($!

'()*+,-*./0
'"!

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"($$

!"#$!

!"#$'

!"#$&

'()*+,-*./1
#23*3/*./0
#23*3/*./1

!"&

#23*+,-*./0

#23*+,-*./1
!"%

!"$

!"#

!"#$%

!"!
$

$)'

$)%

$)*

$)+

!

!

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

!"#

!"$

!"%

!"&

'

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

Figure 4. Steady-state water-oil relative permeability data for Sample 8 in the preserved state. Three hysteresis cycles were

measured: primary imbibition (1st imb), secondary drainage (2nd dr), and secondary imbibition (2nd imb). Steady-state data are
designated by individual symbols (!
! , krow;!, krw), and unsteady-state data are designated by lines.

'"$

!"($!

'"#

'()*+,-*./0

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"($$

!"#$!

!"#$'

'()*+,-*./1
'"!

#23*3/*./0
#23*3/*./1

!"&

#23*+,-*./0
#23*+,-*./1

!"%

!"$

!"#$&

!"#

!"!

!"#$%
$

$)'

$)%

$)*

$)+

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

!

!

!"#

!"$

!"%

!"&

'

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

Figure 5. Steady-state water-oil relative permeability data for Sample 9 in the preserved state. Three hysteresis cycles were
measured: primary imbibition (1st imb), secondary drainage (2nd dr), and secondary imbibition (2nd imb). Steady-state data are
designated by individual symbols (!
! , krow;!, krw), and unsteady-state data are designated by lines.


10

SPE 161932

!"($!

'"!

&"#

$*+,-./,012

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"($$

!"#$!

!"#$'

$*+,-./,013

&"!

%45,51,012
%"#

%45,51,013
%45,-./,012

%"!

%45,-./,013

$"#

$"!

!"#$&
!"#

!"#$%

!"!
$

$)'

$)%

$)*

$)+

!

!

!"%

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

!"'

!"(

!")

$

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

%#

%#

&#

&#

#
#

#'%

#'"

#'(

#')

&

!&#

&*+,-./0,12+-34
!%#

%56,670,12+-34
&*+,-./0,84*+9746

!"#$%%"&'()&*++,&*-(#+$

!"#$%%"&'()&*++,&*-(#+$

Figure 6. Steady-state water-oil relative permeability data for Sample 10 in the preserved state. Three hysteresis cycles were
measured: primary imbibition (1st imb), secondary drainage (2nd dr), and secondary imbibition (2nd imb). Steady-state data are
designated by individual symbols (!
! , krow;!, krw), and unsteady-state data are designated by lines.

#
#

#'"

#'(

#')

&

!&#

&*+,-./0,12+-34
!%#

%56,670,12+-34
&*+,-./0,84*+9746
%56,670,84*+9746

%56,670,84*+9746
!$#

!$#

!"#

#'%

./&0"%$1*2(3"4*&(5"4,&"4$/6-(7&"84$/6

Figure 7. A comparison of primary imbibition and secondary
drainage capillary pressure data for Sample 5 in the
preserved and restored wettability states. Results for Sample
5 are representative of all other samples except for Sample 2.

!"#

./&0"%$1*2(3"4*&(5"4,&"4$/6-(7&"84$/6

Figure 8. A comparison of primary imbibition and secondary
drainage capillary pressure data for Sample 2 in the
preserved and restored wettability states. Results appear
more reproducible. Notably, Sample 2 permeability was an
order of magnitude greater than the other samples.


SPE 161932

11

!"($!

'"#

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

!"($$

!"#$%&'"()"*+"$,&#&%-.(/*$0%&12(31.45&

'()*+,-*./01
23)+45
'()*+,-*./61
23)+45
'()*+,-*./01
75()0/58
'()*+,-*./61
75()0/58

'"!

!"#$!

!"#$'

!"#$&

!"&

!"%

!"$

!"#

!"#$%

!"!
$

$)'

$)%

$)*

$)+

!

!

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

!"#

!"$

!"%

!"&

'

61*+$#&7"8(9$%"*(4$%:*$%&12.(/*$0%&12

Figure 9. A comparison of primary imbibition water-oil relative permeability data for Sample 8 in the preserved and restored
wettability states. Steady-state data are designated by individual symbols (!
! , krow;!, krw), and unsteady-state data are
designated by lines.

Flow Direction
P1
P2

!"#

!"$

P3

54.9 cm

!"%

!"&

P4
Figure 10. Schematic of the whole core composite used in the whole core waterflood. Samples labeled

from top to bottom are WC1, WC2, WC3, and WC4. Sample WC2 contained a visibile stylolitic interval.


12

SPE 161932

Figure 11. Production and pressure drop data from the whole core waterflood.

Figure 12. Water saturation profiles at various points during the whole core waterflood.


SPE 161932

13

3'4'796'88:;#"2'3# 4'=>@77'88

3'4'79A'88:;#"2'3# 4'6=A@7'88

3'4'79?'88:;#"2'3# 4'<5577'88

3'4'796'88:;#"2''3# 4'<=>77'88

,%$-*'.*-%/$0*+)102'3# 4'567'88

3'4'796'88:;#"2''3# 4'==<'88

3'4'796'88:;#"2''3# 4'6<@'88

3'4'796'88:;#"2''3# 4'56'88

!"#$#%&'(%$)*%$#+"

Figure 13. Comparison between whole core experimental and simulation results. Oil production is
normalized to composite pore volume.

Figure 14. Screenshots of water saturation at various times in the core-scale simulation of the whole
core waterflood experiment. Screenshots are identified by injection flowrate and injected brine volume
at the time of the screenshot.


14

SPE 161932

'#

01123
'"
##

!"#$%$"&'()"&*'+",-.'/0

#"

+,-.//

456787,52

&#

"

*$
$!'

&"

%%!

78923

%#

*$"
!%#""

%"

!**""
$%)'"

$#

%#'""

$"
!#
!"

"

"($

"(&

"('

"()

!

12%-,'32%4,2%$"&.'5,2/%$"&

Figure 15. Saturation profiles from the core scale simulation along length
of core at different brine injection volumes.

!"#"$"%&'

!"#"()"%&'
!"#"*+"%&'
!"#"),"%&'
!"#",)"%&'
!"#"(*("%&'
-./"'0!1&0!.-2

Figure 16. Screen captures of oil saturation profies of the mechanistic
model at various times.