Bui Ku 0099M 11098 DATA 1

Near-Miscible CO2 Application to Improve Oil Recovery
By
Ly Huong Bui
Submitted to the graduate degree program in Chemical and Petroleum Engineering and the
Graduate Faculty of the University of Kansas in partial fulfillment of the requirement for
the degree of Master of Science
Committee: ____________________________
G. Paul Willhite
___________________________
Jyun Syung Tsau
___________________________
Aaron M. Scurto
Date defended: __________________________

Thesis Committee for Ly Huong Bui certifies
that this is the approved version of the following thesis
Near-Miscible CO2 Application to Improve Oil Recovery
Committee: ____________________________
G. Paul Willhite
___________________________
Jyun Syung Tsau
___________________________
Aaron M. Scurto
Date approved: __________________________
i
ABSTRACT
Carbon dioxide (CO2) injection for enhanced oil recovery is a proven technology. CO2
injection is normally operated at a pressure above the minimum miscibility pressure
(MMP), which is determined by crude oil composition and reservoir conditions. This is the
lowest pressure at which the injected CO2 becomes dynamically miscible with the crude oil
remaining in the reservoir. However, many reservoirs are located at depths or under
geologic conditions such that they must operate at pressures below the MMP. When CO2 is
injected at below the MMP, displacement efficiency decreases as a result of the loss of
miscibility. CO2 injection is usually not considered as an enhanced oil recovery process in
these reservoirs. Near miscible displacement generally refers to the process that occurs at
pressures slightly below the MMP, but the actual pressure range has never been clearly
defined.
The objectives of this study were to investigate the feasibility of near-miscible CO2
application and improve our understanding of the mechanisms of near-miscible CO2
flooding by conducting appropriate experimental work and reservoir simulation. The
pressure range of interest was from 0.8 MMP to MMP in our study. The Arbuckle formation
of Kansas was used as an example to demonstrate our approach to evaluate CO2 flooding at
near-miscible conditions. The suite of laboratory experiments used to evaluate the
feasibility of operating at pressures below MMP for Arbuckle reservoirs included phase
behavior studies, core flow tests and phase behavior model construction using CMG
software package.
ii
Phase behavior studies were carried out to characterize the near miscible conditions. Slim
tube displacements and swelling/extraction tests were performed to identify the near
miscible range and the mass transfer mechanisms which were responsible for the oil
recovery within this range. A phase behavior model was constructed and well-tuned to
simulate oil properties, CO2/crude oil interactions and slim tube results. Core flow tests
were conducted to evaluate the oil recovery efficiency in the near miscible range.
Initial laboratory works indicated that miscibility was not achievable, however at least
65% to 80% of the waterflood residual oil for dolomite cores and lesser from 45% to 60 %
for sandstone core in the near-miscible region was observed. The principal oil recovery
mechanism in the near-miscible range appeared to be extraction/vaporization of
hydrocarbon components from crude oil into the CO2 rich vapor phase, coupled with
enhanced mobility control due to the reduction of oil viscosity. This suggested that
application of carbon dioxide in the field would require injection and recycling of large
volumes of carbon dioxide. Further study is needed to determine if such a process is
economically feasible. However the prospect of recovering up to 1 billion barrels of oil from
Arbuckle reservoirs offers significant economic potential.
iii
TABLE OF CONTENTS
ABSTRACT ................................................................................................................................................ II
TABLE OF CONTENTS .......................................................................................................................... IV
LIST OF TABLES..................................................................................................................................... XI
NOMENCLATURE ................................................................................................................................ XII
ACKNOWLEDGMENTS ...................................................................................................................... XIV
1 INTRODUCTION AND LITERATURE REVIEW ....................................................................... 1
1.1 THE BASICS OF CO2 EOR.......................................................................................................................... 4
1.1.1 CO2 Properties......................................................................................................................................... 4
1.1.2 Immiscibility & Miscibility................................................................................................................. 6
1.1.3 Minimum Miscibility Pressure ......................................................................................................... 8
1.1.4 Mechanisms of Oil Displacement by CO2 ..................................................................................... 9
1.1.5 Mechanisms for CO2 Miscibility with Oil .................................................................................. 11
1.2 MOTIVATIONS BEHIND THIS STUDY....................................................................................................... 15
1.3 OBJECTIVES OF THE STUDY ..................................................................................................................... 21
2 PHASE BEHAVIOR STUDIES ..................................................................................................... 23
2.1 FLUID PROPERTIES .................................................................................................................................. 23
2.2 SLIM TUBE DISPLACEMENTS ................................................................................................................. 26
2.2.1 Experimental Setup and Specifications .................................................................................... 27
2.2.2 Experimental Procedures ............................................................................................................... 29
2.2.3 Results and Discussions ................................................................................................................... 31
iv
2.2.4 Conclusions ........................................................................................................................................... 36
2.3 SWELLING/EXTRACTION TESTS ............................................................................................................ 37
2.3.1 Experimental Setup and specifications .................................................................................... 37
2.3.3 Experimental Principles .................................................................................................................. 41
2.3.4 Apparatus Validation ....................................................................................................................... 43
2.3.5.1 Effect of System Pressure ........................................................................................................................ 44
2.3.5.2 Effect of System Temperature ................................................................................................................. 47
2.3.5.3 Effect of Initial Oil Volume ....................................................................................................................... 50
2.3.6 Conclusions ........................................................................................................................................... 55
2.4 VISCOSITY MEASUREMENTS................................................................................................................... 55
2.4.1 Principle of Operation ...................................................................................................................... 56
2.4.2 Experimental Setup and Specifications .................................................................................... 57
2.4.3.1 Oil Viscosity Measurement Procedure .................................................................................................... 59
2.4.3.2 Oil/CO2 Mixture Viscosity Measurement Procedure ............................................................................. 61
2.4.5 Conclusions ........................................................................................................................................... 63
2.5 PHASE BEHAVIOR MODEL ...................................................................................................................... 63
2.5.1 Phase Behavior Modeling using WINPROP ............................................................................ 65
2.5.2 Equation of State Characterization ........................................................................................... 68
2.5.3 Slim Tube Modeling using GEM ................................................................................................... 70
3 CORE FLOW TESTS ..................................................................................................................... 76
v
3.1 CORES ........................................................................................................................................................ 76
3.2 FLUIDS ....................................................................................................................................................... 79
3.3 EQUIPMENTS AND PROCEDURES ........................................................................................................... 79
3.3.1 Core Characterization ..................................................................................................................... 79
3.3.1.1 Pore Volume Measurements.................................................................................................................... 80
3.3.1.1.1 Gravimetric Method ............................................................................................................................ 80
3.3.1.1.2 Tracer Tests ......................................................................................................................................... 81
3.3.1.2 Permeability Measurements.................................................................................................................... 85
3.3.2 Core Floods ........................................................................................................................................... 86
3.3.2.1 Experimental Setup .................................................................................................................................. 86
3.3.2.2 Experimental Procedures ........................................................................................................................ 88
3.3.2.3 Displacement Rate Selection ................................................................................................................... 88
3.4 RESULTS AND DISCUSSIONS ................................................................................................................... 89
3.4.1 Core Characterization Results...................................................................................................... 89
3.4.2 Core Floods Results ........................................................................................................................... 89
3.4.2.1 Secondary CO2 Flooding........................................................................................................................... 89
3.4.2.2 Tertiary CO2 flooding ............................................................................................................................... 91
3.4.2.3 Effect of Water Saturation on Oil Recovery Efficiency........................................................................... 95
3.5 CONCLUSIONS ........................................................................................................................................... 99
4 CONCLUSIONS & RECOMMENDATIONS ............................................................................ 101
4.1 CONCLUSIONS ......................................................................................................................................... 101
4.2 RECOMMENDATIONS ............................................................................................................................. 102
REFERENCES....................................................................................................................................... 104
APPENDICES ....................................................................................................................................... 107
vi
LIST OF FIGURES
FIGURE 1-1 GAS INJECTION EOR IN U.S [1] ............................................................................................................. 2
FIGURE 1-2 MISCIBLE CO2 GAS INJECTION EOR IN US [1] ................................................................................... 3
FIGURE 1-3 COMPARISON OF CO2, CH4, N2 DENSITY AT 110OF (REFPROP) .................................................. 5
FIGURE 1-4 COMPARISON OF CO2, CH4, N2 VISCOSITY AT 110OF (REFPROP) ............................................... 6
FIGURE 1-5 MISCIBILITY OF PROPANE (OR LPG) LIQUID AND OIL LIQUID AT RESERVOIR TEMPERATURE AND
PRESSURE CONDITIONS [3] ................................................................................................................................. 7
FIGURE 1-6 IMMISCIBLE TWO-PHASE MIXTURE OF METHANE GAS AND OIL LIQUID AT TYPICAL RESERVOIR
CONDITIONS [3] .................................................................................................................................................... 7
FIGURE 1-7 EFFECT OF TEMPERATURE AND PRESSURE ON CO2 INJECTION DISPLACEMENT MECHANISMS [6]
................................................................................................................................................................................ 9
FIGURE 1-8 ONE DIMENSIONAL SCHEMATIC SHOWING HOW CO2 BECOMES MISCIBLE WITH CRUDE OIL ...... 11
FIGURE 1-9 CONCEPT OF MULTIPLE-CONTACT MISCIBILITY BY VAPORIZATION [3] ......................................... 12
FIGURE 1-10 CONCEPT OF MULTIPLE CONTACT MISCIBILITY BY CONDENSATION [3] ..................................... 14
FIGURE 1-11 OIL PRODUCTION FROM ARBUCKLE FORMATION IN TOTAL KANSAS OIL PRODUCTION ............ 15
FIGURE 1-12 EFFECT OF PRESSURE AND PORE VOLUME INJECTED ON TERTIARY FLOOD OIL RECOVERIES IN
SYSTEM 1: CO2/DECANE/BEREA [7] ........................................................................................................... 17
FIGURE 1-13 EFFECT OF PRESSURE AND PORE VOLUME INJECTED ON TERTIARY OIL RECOVERIES IN SYSTEM
2: CO2/WEST TEXAS CRUDE/MIXED-WET CARBONATE CORE AT 100OF [7] ........................................ 17
FIGURE 1-14 COMPARISON OF RESERVOIR S CO2 COREFLOODS WITH TEXAS CREAM LIMESTONE CORE AND
SLIM-TUBE TESTS [11] ..................................................................................................................................... 19
vii
FIGURE 1-15 COMPARISON OF RESERVOIR H CO2 COREFLOODS WITH TEXAS CREAM LIMESTONE CORE AND
SLIM-TUBE TESTS [11] ..................................................................................................................................... 19
FIGURE 1-16 COMPARISON OF RESERVOIR M CO2 COREFLOODS WITH TEXAS CREAM LIMESTONE CORE AND
SLIM-TUBE TESTS [11] ..................................................................................................................................... 20
FIGURE 1-17 COMPARISON OF HYDROCARBON LEAN GAS SLIM TUBE TESTS AND RESERVOIR A COREFLOODS
[11]..................................................................................................................................................................... 20
FIGURE 1-18 GENERALIZED RECOVERY RESPONSE TO PRESSURE ....................................................................... 22
FIGURE 2-1 OGALLAH UNIT, TREGO COUNTY, KANSAS ........................................................................................ 24
FIGURE 2-2 GC COMPOSITIONAL ANALYSIS RESULT OF OGALLAH CRUDE OIL ................................................... 25
FIGURE 2-3 SCHEMATIC OF SLIM TUBE SETUP........................................................................................................ 27
FIGURE 2-4 RESULTS OF DISPLACEMENT TESTS AT 110OF.................................................................................. 31
FIGURE 2-5 RESULTS OF DISPLACEMENT TESTS AT 125OF.................................................................................. 32
FIGURE 2-6 MINIMUM MISCIBILITY PRESSURE DETERMINATIONS AT 110OF AND 125OF ............................. 33
FIGURE 2-7 NEAR MISCIBLE REGION FOR OGALLAH CRUDE OIL AT 110OF ....................................................... 34
FIGURE 2-8 DENSITY PROFILE OF THE EFFLUENT AT 110OF............................................................................... 35
FIGURE 2-9 DENSITY PROFILE OF THE EFFLUENT AT 125OF............................................................................... 36
FIGURE 2-10 EXPERIMENTAL SETUP INCLUDE (1) GAS CYLINDER (2) ISCO SYRINGE PUMP (3) FISHER
ISOTEMP CIRCULATOR (4) FISHER ISOTEMP IMMERSION CIRCULATOR (5) WATER BATH (6) HIGH
PRESSURE VIEW CELL (7) MIXING BAR (8) LABORATORY JACK (9) COMPUTER (10) CATHETOMETER
WITH TELESCOPE (11) VACUUM PUMP [13]................................................................................................. 37
FIGURE 2-11 AN ACTUAL IMAGE OF THE SWELLING/EXTRACTION EXPERIMENTAL SETUP............................. 39
viii
FIGURE 2-12 COMPARISON OF LIQUID PHASE COMPOSITIONS FOR CO2+ N-DECANE SYSTEM AT 71.10OC
(160OF) WITH LITERATURE DATA (○) THIS WORK (□) THIS WORK (∆) NAGARAJAN & ROBINSON JR.
(◊) JENNINGS & SCHUCKER ............................................................................................................................. 44
FIGURE 2-13 CHANGE OF INITIAL OIL VOLUME WITH PRESSURE ........................................................................ 45
FIGURE 2-14 EFFECT OF PRESSURE ON CO2 SOLUBILITY AND SWELLING FACTOR AT 110OF ........................ 46
FIGURE 2-15 EFFECT OF TEMPERATURE ON CO2 SOLUBILITY ............................................................................ 48
FIGURE 2-16 EFFECT OF TEMPERATURE ON SWELLING/ EXTRACTION CURVES .............................................. 48
FIGURE 2-17 DEPENDENCE OF CO2 DENSITY ON TEMPERATURE AND PRESSURE (REFPROP) ................... 50
FIGURE 2-18 EFFECT OF INITIAL OIL VOLUME ON CO2 SOLUBILITY AT 110OF ................................................ 51
FIGURE 2-19 EFFECT OF INITIAL OIL VOLUME ON SWELLING/EXTRACTION CURVE AT 110OF ..................... 51
FIGURE 2-20 COMPUTED P-X DIAGRAM FOR THE FLOUNDER OIL/CO2 SYSTEM AT 233OF [16] .................. 53
FIGURE 2-21 COMPUTED P-X DIAGRAM FOR OGALLAH OIL/ CO2 SYSTEM AT 110OF ..................................... 54
FIGURE 2-22 SCHEMATIC OF HIGH-PRESSURE VISCOSITY MEASUREMENT SETUP ............................................ 57
FIGURE 2-23 AN ACTUAL IMAGE OF THE HIGH-PRESSURE VISCOSITY MEASUREMENT SETUP ......................... 58
FIGURE 2-24 A HIGH PRESSURE GENERATOR ......................................................................................................... 59
FIGURE 2-25 EFFECT OF PRESSURE ON OGALLAH VISCOSITY AT 110OF ........................................................... 62
FIGURE 2-26 EFFECT OF CO2 DISSOLUTION IN CRUDE OIL ON THE VISCOSITY OF OGALLAH FLUID ............... 62
FIGURE 2-27 COMPARISON OF VISCOSITY/DENSITY EXPERIMENTAL DATA AND SIMULATED DATA
BEFORE/AFTER EOS TUNING .......................................................................................................................... 69
FIGURE 2-28 COMPARISON OF SATURATION PRESSURE/SWELLING FACTOR EXPERIMENTAL DATA AND
SIMULATED DATA BEFORE/AFTER EOS TUNING .......................................................................................... 70
FIGURE 2-29 RELATIVE PERMEABILITY CURVE OF OIL-GAS USED IN THE SIMULATION [18].......................... 72
FIGURE 2-30 COMPARISON OF SIMULATED MMP AND EXPERIMENTAL MMP ................................................ 73
ix
FIGURE 2-31 RELATIVE PERMEABILITY CURVE OF OIL-GAS BEFORE AND AFTER ADJUSTED TO MATCH SLIM
TUBE RESULTS .................................................................................................................................................... 74
FIGURE 2-32 COMPARISON OF SIMULATED MMP AND EXPERIMENTAL MMP AFTER TUNING THE RELATIVE
PERMEABILITY CURVE ....................................................................................................................................... 75
FIGURE 3-1. VACUUM AND SATURATION SETUP..................................................................................................... 78
FIGURE 3-2 EQUAL AREA TECHNIQUE IN CALCULATING PV OF A CORE FROM TRACER TESTS ......................... 82
FIGURE 3-3 SCHEMATIC OF TRACER SETUP ............................................................................................................ 84
FIGURE 3-4 SCHEMATIC OF CORE FLOOD SETUP .................................................................................................... 86
FIGURE 3-5 PERCENT OF OIL RECOVERY @ 6 PV OF CO2 INJECTED IN CORE FLOW TESTS AND PERCENT OF
OIL RECOVERY @ 1.2 PV OF CO2 INJECTED IN SLIM TUBE TESTS AT 110OF ........................................... 90
FIGURE 3-6 PERCENT OF OIL RECOVERY @ 6 PV OF CO2 INJECTED AS A FUNCTION OF PRESSURE FOR BEREA
SANDSTONE A1, BAKER DOLOMITE AND ARBUCKLE DOLOMITE AT 110OF ............................................. 93
FIGURE 3-7 RECOVERY EFFICIENCY OF BEREA SANDSTONES A1 & A2 AT 1100 PSIG & 110OF. ................. 95
FIGURE 3-8 EFFECT OF INITIAL WATER SATURATION ON CO2 FLOOD OIL RECOVERY PERFORMANCE BY
COMPARING SECONDARY AND TERTIARY CO2 FLOODS [7] .......................................................................... 97
FIGURE 3-9 EFFECT OF WATER SATURATION ON OIL RECOVERY EFFICIENCY AT 110OF ................................. 98
FIGURE 3-10 CO2 FLOOD END-POINT PERMEABILITY AS A FUNCTION OF WATER SATURATION AND RESIDUAL
OIL SATURATION IN A MIXED WET CARBONATE [7] ...................................................................................... 99
x
LIST OF TABLES
TABLE 2.1 PHYSICAL PROPERTIES OF OGALLAH CRUDE OIL AND THE LUMPED HEAVY COMPONENT C36+ . 25
TABLE 2.2 SLIM TUBE PROPERTIES.......................................................................................................................... 28
TABLE 2.3 PHASE EQUILIBRIUM DATA OF CO2/N-DECANE AT 71.1OC.............................................................. 43
TABLE 2.5 ADJUSTMENTS OF EOS PARAMETERS .................................................................................................. 68
TABLE 2.6 MAXIMUM PERCENTAGE ERROR BETWEEN SIMULATED AND EXPERIMENTAL DATA .................... 70
TABLE 2.4 SLIM TUBE MODEL PROPERTIES ............................................................................................................ 71
TABLE 3.1 DENSITY/VISCOSITY OF BRINE AT 86OF/110OF .............................................................................. 79
TABLE 3.2 CORE PROPERTIES ................................................................................................................................... 89
TABLE 3.3 TERTIARY CO2 FLOOD RESULTS IN OGALLAH/BEREA SANDSTONE A1 AT 110OF ....................... 92
TABLE 3.4 TERTIARY CO2 FLOOD RESULTS IN OGALLAH/BEREA SANDSTONE A2 AT 110OF ....................... 92
TABLE 3.5 TERTIARY CO2 FLOOD RESULTS IN OGALLAH/ARBUCKLE DOLOMITE AT 110OF ......................... 92
TABLE 3.6 TERTIARY CO2 FLOOD RESULTS IN OGALLAH/ BAKER DOLOMITE AT 110OF ............................... 92
xi
NOMENCLATURE
Swr Connate water saturation
Sorw Residual oil saturation after waterflooding
Sorm Residual oil saturation after CO2 flooding
Swf Water saturation after CO2 flooding
krg Gas relative permeability
kro Oil relative permeability
Sg Gas saturation
µ Viscosity
Tc Critical temperature
T Temperature
Pc Critical pressure
P Pressure
R Molar gas constant
M Molecular weight
Ωa Equation of state parameter
Ωb Equation of state parameter
ω Accentric factor
dij Binary interaction coefficient
α Rotational coupling coefficient
xCO2 Mole fraction of CO2 in the liquid phase
xii
MW Molecular Weight
PVT Pressure Volume Temperature
PV Pore volume
HCPV Hydrocarbon pore volume
xiii
ACKNOWLEDGMENTS
First and foremost, I would like to express my sincere gratitude to my advisors, Dr. G. Paul
Willhite and Dr. Jyun-Syung Tsau for their guidance, support and encouragement
throughout my graduate study at University of Kansas. I appreciate their patience to guide
me through and make graduate study an invaluable experience to me.
I would like to extend my appreciation to Dr. Aaron M. Scurto for serving on my thesis
committee and giving me insightful comments.
I would like to give my special thanks to Mr. Scott Ramskill and Dr. Karen Peltier for their
assistance on setting up and maintaining laboratory equipment as well as providing me
necessary laboratory supplies to get my work done.
I would like to thank the faculty and staff members of the University of Kansas Center for
Research’s Tertiary Oil Recovery Project (TORP), especially Dr. Jenn-Tai Liang and Ms.
Mayumi Crider. I would also like to thank other fellow students of the Chemical and
Petroleum Department for their friendship.
I would like to acknowledge Research Partnership to Secure Energy for America (RPSEA)
and the University of Kansas Center for Research’s Tertiary Oil Recovery Project (TORP)
for financial support. I would like to acknowledge Kansas Geological Survey and Carmen
Schmitt, Inc. for providing representative core samples and crude oil samples. I would like
xiv
to acknowledge Computer Modeling Group (CMG) for their permission of using the
software.
Finally I want to express the deepest appreciation to my mother and my sister for their love
and support.
xv
1 Introduction and Literature Review
Crude oil production can include up to three phases: primary, secondary, and tertiary (or
enhanced) oil recovery. Primary oil recovery, which relies on natural reservoir pressure to
drive the oil to the surface, typically produces 5 to 20 % of the original oil in place (OOIP).
Secondary recovery techniques prolong the field's productive life generally by injecting
water or gas to displace oil to production wells, increasing the oil recovery from 20 to 40%
OOIP. Several tertiary, or enhanced oil recovery (EOR) techniques, which have been
attempted with the goal of ultimately recovering/producing 30 to 60 % OOIP and found to
be commercially successful to varying degrees, could be categorized mainly as thermal
recovery, gas injection and chemical injection.
Gas injection involves injecting natural gas, nitrogen or carbon dioxide (CO2) into the
reservoir. The gases can either expand and push oil through the reservoir, or dissolve in the
oil, decreasing its viscosity and facilitating oil flow to the wellbore. While other EOR
methods decline over the years, gas injection increased from 18% in 1984 to about 48% in
2008, proving to be the most popular method in the U.S. The percentage of gas injection
share in total EOR production in the U.S is depicted in Figure 1.1.
1
Figure 1-1 Gas injection EOR in U.S [1]
CO2 injection for secondary and tertiary oil recovery is the primary gas injection method.
Miscible CO2 displacement offers the greatest oil recovery potential, but it can only be
achieved at a pressure greater than a certain minimum referred to as the minimum
miscibility pressure. In a study of EOR developments and their future potential in the U.S,
Stosur et al. [1] concluded that miscible CO2 gas injection was slowly becoming popular and
would continue to grow faster than any other EOR methods. He pointed out that miscible
CO2 gas injection increased from 38% in 1984 to about 65% in 2004.
2
Figure 1-2 Miscible CO2 gas injection EOR in US [1]
CO2 EOR not only provides a value-added opportunity to recover substantial remaining
trapped oil but also offers environmental benefits if industrial CO2 is used and stored in
reservoirs. With existing but non-utilized industrial CO2 sources, such as natural gas
byproducts, hydrogen plants, ethanol, cement, petroleum coke gasifiers, fossil fuel power
plants, the growing concern about greenhouse gases and climate change has significantly
increased interest in CO2 sequestration and storage.
3
1.1 The Basics of CO2 EOR
1.1.1 CO2 Properties
CO2 has numerous characteristics that make it a favorable oil-displacing-agent. One of the
most important characteristics is the pressure at which CO2 becomes miscible with crude
oil is substantially lower than other gases, although there may be exceptions at high
temperature. When miscibility is attained, residual oil saturation is reduced nearly to zero,
which means high oil recoveries and favorable project economics.
High solubility of CO2 in crude oil is another important characteristic. CO2 dissolution in oil
causes the oil to swell and expand out of dead end pores. In addition, viscosity of crude oil
decreases as it becomes saturated with CO2 at increasing pressures, which helps facilitating
flow to the wellbore. The ability of dense-phase CO2 to extract hydrocarbon components
from oil, coupled with the ability of CO2 to dissolve into oil helps establishing dynamic
miscibility.
At high pressures, CO2 density has a density close to that of a liquid and is greater than that
of either nitrogen (N2) or methane (CH4), which makes CO2 less prone to gravity
segregation compared with N2 or CH4.
4
0.9
CH4
0.8 CO2
0.7 N2
0.6
Density, g/cc
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000
Pressure, psi
Figure 1-3 Comparison of CO2, CH4, N2 density at 110oF (REFPROP)
In addition, at high pressures, viscosity of CO2 is also greater than that of N2 or CH4,
resulting in better mobility control and better sweep efficiency compared with other gases.
5
0.09
CH4
0.08
CO2
0.07 N2
0.06
0.05
Viscosity, cp
0.04
0.03
0.02
0.01
0
0 500 1000 1500 2000 2500 3000
Pressure, psi
Figure 1-4 Comparison of CO2, CH4, N2 viscosity at 110oF (REFPROP)
1.1.2 Immiscibility & Miscibility
Holm [2] describes miscibility as: “the ability of two or more substances to form a single
homogeneous phase when mixed in all proportions. For petroleum reservoirs, miscibility is
defined as that physical condition between two or more fluids that will permit them to mix
in all proportions without existence of an interface. If two fluid phases form after some
amount of one fluid is added to others, the fluids are considered immiscible”. Examples of
miscibility and immiscibility are shown in Figure 1.5 and Figure 1.6 respectively.
6
Figure 1-5 Miscibility of propane (or LPG) liquid and oil liquid at reservoir temperature and pressure
conditions [3]
Oil exists as a liquid and propane exists as a gas at atmospheric conditions. If the
temperature and pressure are increased to the indicated reservoir conditions, however, the
propane exists as a liquid, as does the crude oil. These two liquids will mix completely and
are thus miscible.
Figure 1-6 Immiscible two-phase mixture of methane gas and oil liquid at typical reservoir conditions [3]
Methane and crude oil are partially soluble in one another. At conditions typically
encountered in a reservoir, however, they do not mix in all proportions and two phases
exist.
7
1.1.3 Minimum Miscibility Pressure
Miscible recovery of reservoir oil can be achieved by CO2 displacement at a pressure
greater than a certain minimum. This minimum pressure is hereafter called as the CO2
minimum miscibility pressure (MMP). CO2 MMP is an important parameter for screening
and selecting reservoirs for CO2 EOR. For the highest recovery, a candidate reservoir must
be capable of withstanding an average reservoir pressure greater than the CO2 MMP.
MMP depends on crude oil composition and reservoir conditions, and is typically
determined using slim tube tests. There are no fixed criteria for determining miscibility
within slim tube and individual researchers have defined their own criteria to identify slim
tube miscibility. Holm and Josendal [4] defined MMP as the pressure at which 80% of oil in
place was recovered at CO2 breakthrough and more than 94% when gas-oil ratio (GOR)
reaches 40,000 scf/bbl, whereas Metcalfe and Yellig [5] defined MMP as a pressure at
which recovery at 1.2 pore volume gas injected was near the maximum recovery. Others
used the pressure corresponding to the break-over point in the plot of percentage of oil
recovery at 1.2 HCPV of CO2 injected versus pressure as the MMP. In this study, the
pressure at which 90% of original oil in place was recovered at 1.2 HCPV of CO2 injected
was considered MMP.
8
1.1.4 Mechanisms of Oil Displacement by CO2
Mechanisms of CO2 displacing crude oil from porous media rely on the phase behavior of
CO2 /crude oil system, which is strongly dependent on reservoir temperature, pressure and
crude oil composition. Displacement mechanisms fall into one of five regions illustrated in
Figure 1.7. It is important to note that the lines that divide region from region are
generalizations that will vary from oil to oil.
Figure 1-7 Effect of temperature and pressure on CO2 injection displacement mechanisms [6]
In Region I (low pressure applications), CO2 swells the oil, reduces the viscosity of crude oil
and contributes to internal solution gas drive. Swelling of oil is important since the oil left
in the reservoir after flooding is inversely proportional to the swelling factor. Swollen oil
droplets will force water out of pore spaces, creating drainage rather than imbibitions
9
process for water wet system. Drainage oil relative permeability curves are higher than the
imbibitions counterparts, creating a more favorable oil flow environment at any given
saturation conditions. The importance of viscosity reduction was mentioned earlier.
Another mechanism of oil displacement by CO2 in Region I is solution gas drive effect. CO2
goes into the solution with an increase in reservoir pressure, after termination of the
injection phase of a gas flood, gas will come out of solution and continue to drive oil to the
wellbore.
At reservoir pressures higher than those in Region I but lower than Region IV,
supplemental production mechanisms come into play in Region II. In addition to increasing
reservoir pressure, oil swelling, viscosity reduction, hydrocarbons may be vaporized into
the gas phase.
Region III (intermediate pressure, low temperature applications) is very similar to Region
II, except that CO2, rather than vaporize crude oil, extract the crude’s lighter hydrocarbons
forming CO2-rich liquid mixtures. In addition to the potential for reduced total mobility in
the three-phase region, Orr et al (1983) report that CO2-rich liquid phases extract more and
much heavier hydrocarbons than their rich vapor counterparts.
Region IV is the most important region where CO2 vaporizes or extracts a significant
amount of hydrocarbon components from crude oil so rapidly that multiple-contact
miscibility occurs in a very brief time period and over a very short reservoir distance.
10
1.1.5 Mechanisms for CO2 Miscibility with Oil
In general, miscibility between two fluids can be achieved through two mechanisms: first
contact miscibility and multiple contact miscibility.
Dynamic Miscibility
Oil
CO2
CO2 CO2 Original

Pure
vaporizing oil condensing Oil
CO2
components into oil
Direction of displacement
Figure 1-8 One dimensional schematic showing how CO2 becomes miscible with crude oil
CO2 is not first contact miscible with most crude oils within the reasonable range of
reservoir pressures. It requires many contacts in which components of the oil and CO2
transfer back and forth until the oil-enriched CO2 cannot be distinguished from the CO2-
enriched oil in terms of fluid properties. These processes are called vaporizing/condensing
processes and are described in detail in Green & Willhite [3].
Ternary diagram is used to describe the phase behavior of three-component systems at
constant system temperature and pressure and is useful in describing the development of
miscibility in a multiple contact process. The apexes of the diagram represent 100%
11
concentration of methane, nitrogen or CO2; intermediate hydrocarbons (C2-C6) and heavy
hydrocarbons (C7+) respectively. Temperature and pressure are constant and equal to the
displacement conditions in the reservoir. VO is the saturated vapor curve. LO is the
saturated liquid curve. Point O is the critical or plait point. Point A represents the
composition of injected gas A. Point C represents the composition of crude oil. The area
inside the envelope is the two-phase region and outside the envelope is the single phase
region.
Figure 1-9 Concept of multiple-contact miscibility by vaporization [3]
Concept of multiple-contact miscibility by vaporizing is illustrated in Figure 1.9. The
injected gas is a relatively lean gas, for example, it contains mostly methane and other low
molecular weight hydrocarbons or sometimes inert gases such as nitrogen. In this
approach, the composition of the injected gas is modified as it moves through the reservoir
so that it becomes miscible with the original reservoir oil, that is, the injected fluid is
enriched in composition through multiple contacts with the oil, during which intermediate
12
components in the oil are vaporized into the injected gas. The process operates
conceptually as follows:
 Gas A mixes with Oil C. The resulting composition of the mixture is along AC, say
Point a.
 Mixture a is in the two-phase region, therefore, separates into a vapor V1 and a
liquid L1
 Vapor V1 moves ahead of Liquid L1 and contacts Oil C. The resulting composition of
the mixture is along Line V1C, say at Point b.
 Mixture b separates into Vapor V2 and Liquid L2
 The process continues with vapor-phase composition changing along the saturated
vapor curve, V3, V4 etc
 Finally, at point e, the vapor becomes miscible with Oil C because the mixing line lies
in the single-phase region.
If miscibility is lost due to reservoir mixing, miscibility is re-developed the same way as
described above. Also, in order for multiple contact miscibility vaporization process to be
successful, the reservoir fluid composition must lie to the right of the critical tie line. If both
fluids lie to the left, vaporization will occur but not in sufficient quantities to develop
miscibility. If the injection fluids and original oil both lie to the right of the critical tie line,
the fluids are miscible on first contact.
13
Figure 1-10 Concept of multiple contact miscibility by condensation [3]
Concept of multiple-contact miscibility by condensation is illustrated in Figure 1.10. The
injected fluid generally contains larger amounts of intermediate molecular weight
hydrocarbons. In this approach, reservoir oil near the injection well is enriched in
composition by contact with the injected fluid since hydrocarbon components are
condensed from the injected fluid into the oil. Under proper conditions, the oil will be
sufficiently modified in composition to become miscible with additional injected fluid.
Conceptually, the process is very similar to vaporizing process, except vapor phase V1, V2
moves ahead, leaving immobile liquid phase L1, L2 etc to mix with additional injected Gas
A.
For multiple contact miscibility by condensation to occur, the injected fluid must be to the
right of the critical tie line. If it is not, condensation of CO2 into crude oil will still occur,
however, miscibility will not be developed.
14
1.2 Motivations behind this study
The Arbuckle formation has played an important role in total Kansas oil production. These
Arbuckle reservoirs have produced 2.19 billion barrel of oil by a bottom-water drive
mechanism via an underlying aquifer, representing 36% of total Kansas oil production to
date. Today, more than 90% of wells produce less than 5 barrel of oil per day.
Simpson, 4% Viola, 5% Lansing-

Kansas City,
Cherokee, 3% 19%
Marmaton,
3% Mississipian,
16%
Marrow, 3%
Arbuckle
36% Others, 11%
Figure 1-11 Oil production from Arbuckle formation in total Kansas oil production
Previous assessments of CO2 miscible flooding in these Arbuckle reservoirs indicate that
miscibility is not achievable at the current reservoir operating pressures. Therefore, the
possibility of operating at pressures below the MMP means that these reservoirs, which
might be otherwise abandoned with substantial remaining oil left in place, could be
considered for CO2 injection.
15
In addition, a lower-pressure process is attractive from both economic and operational
standpoints, including purchasing smaller gas volumes and decreased gas compression
costs.
In general, pressures below MMP are not high enough to vaporize sufficient oil into the CO2
phase or to allow sufficient CO2 to dissolve into the oil so that the two phases to become
miscible. Loss of miscibility results in loss of oil recovery efficiency. Slim tube results
usually show a dramatic loss of recovery at pressures below MMP. To date, however,
conflicting experimental results have led to considerable disagreement in the literature
regarding the feasibility of operating at pressures below MMP.
Shyeh-Yung et al. [7] studied the effect of operating pressure by conducting tertiary CO2
displacements on two systems,
 System 1-Berea sandstone (strongly water wet)/CO2/decane at 160oF
 System 2-San Andres outcrop carbonate (mixed wet)/CO2/ degassed west Texas
separator oil 100oF
Laboratory results presented in Figure 1.12 and Figure 1.13 showed that tertiary CO2 flood
oil recovery decreased linearly as pressure decreased. No dramatic loss of recovery was
observed below the MMP as suggested by slim tube test results.
16
Figure 1-12 Effect of pressure and Pore Volume Injected on tertiary flood oil recoveries in System 1:
CO2/Decane/Berea [7]
Figure 1-13 Effect of pressure and Pore Volume Injected on tertiary oil recoveries in System 2: CO2/West
Texas crude/Mixed-wet carbonate core at 100oF [7]
17
The author attributed the high oil recoveries at pressures below MMP to the possible
improvement of mobility ratio between CO2 and oil, the low IFT displacement and mass
transfer.
Schechter et al. [8] conducted coreflood tests and simulations to examine the possibility of
optimizing the performance of Wellman Unit by reducing CO2 injection pressure, thereby
reducing the volume of CO2. They found that dropping the pressure from above the MMP to
near the MMP or below the MMP did not reduce the efficiency in laboratory coreflooding.
Similarly, a laboratory study on near-miscible CO2 injection in Steelman Reservoir by Dong
et al. [9] showed that the microscopic displacement efficiency improved with operating
pressure in the near-miscible region, but no dramatic change in oil recovery was observed
with a change in operating pressure.
In a recent overview of industrial experience with CO2 injection, Hadlow [10] cited field
data with good recovery at pressures below the CO2 flood MMP in the North Cross and
Dollarhide reservoirs while they were being pressured up.
In another effort by Grigg et al. [11]; however, a rapid decrease in oil recovery was noted as
pressure fell below the MPMP for the CO2 floods conducted at near the CO2 critical
temperature, as slim tube data showed. He also concluded that the concept of hydrocarbon
gas injection is a more feasible concept, and this maybe the case for CO2 injection at
temperatures well above its critical temperature since the hydrocarbon floods did not
18
show a drastic change in oil recovery near the MMP, unlike the CO2 floods. The residual oil
saturations to gasflood as percent of oil after the waterflood are plotted vs. the ratio of the
flood pressure to the slim tube MMP in Figure 1.14 to Figure 1.17.
Figure 1-14 Comparison of Reservoir S CO2 corefloods with Texas Cream limestone core and slim-tube tests
[11]
Figure 1-15 Comparison of Reservoir H CO2 corefloods with Texas Cream limestone core and slim-tube tests
[11]
19
Figure 1-16 Comparison of Reservoir M CO2 corefloods with Texas Cream limestone core and slim-tube tests
[11]
Figure 1-17 Comparison of hydrocarbon lean gas slim tube tests and Reservoir A corefloods [11]
Finally, the most important motivation behind the study is the environmental benefit of
capturing and sequestering CO2 emissions by utilizing an industrial source of CO2. As
significant volumes of injected CO2 used in the displacement stay in the reservoir formation
20
and the produced CO2 from the productions wells is recycled and re-injected into the
reservoir, CO2 emission into the atmosphere and its impact to the environment is very
limit.
1.3 Objectives of the study
The objectives of this study are to (1) investigate the feasibility of near-miscible CO2
application (2) improve our understanding of the mechanisms of near-miscible CO2
flooding by conducting appropriate experimental work and reservoir simulation. The
Arbuckle formation of Kansas is used as an example to demonstrate our approach to
evaluate CO2 flooding at near-miscible conditions. The suite of laboratory experiments used
to evaluate the feasibility of operating at pressures below MMP for Arbuckle reservoirs
includes phase behavior studies, core flow tests and phase behavior model construction
using CMG software package.
The term near miscible is understood as the transition from immiscible to miscible. Near
miscible refers to displacements at pressures slightly below MMP, where the recovery
efficiency is improved over immiscible displacements. Since it is not clearly defined in the
literature, the pressure range of interest in this study is from 0.8 MMP to MMP.
21
1.0
0.9
0.8
Recovery Efficiency
Region 1
0.7 Miscible
0.6
0.5
0.4
0.3
0.2
Region 2 Region 3
0.1 Immiscible Near Miscible
0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Relative Miscible Pressure, Pres/MMP
Figure 1-18 Generalized recovery response to pressure
22
2 Phase Behavior Studies
Phase behaviors studies are conducted for several purposes: (1) to characterize the oil/CO2
system (2) to understand the mechanism of oil displacement by CO2 (3) to provide
experimental PVT data to fine-tune the phase behavior model.
Phase behavior studies included slim tube displacements, swelling/extraction tests and
viscosity/density measurements. Slim-tube displacements were conducted to determine
the minimum miscibility pressure for this system and to define the near miscible range.
Swelling/extraction tests were performed to examine the mechanisms of oil recovery in the
near miscible region. Viscosity of oil saturated with CO2 was measured at various
pressures. A phase behavior model based on the Peng-Robinson Equation of State (EOS)
was constructed and well-tuned to characterize the fluid properties and the phase behavior
interaction between CO2 and the oil. Slim tube model was constructed using 1-D
compositional simulation with tuned equation of state. Detail descriptions of phase
behavior studies are presented and discussed in this chapter.
2.1 Fluid Properties
All experiments were performed with centrifuged and filtered Ogallah stock tank oil. The
crude oil was centrifuged at the rate of 20 rpm for at least an hour to separate water, oil
and solid particles. Two layers of glass microfibre filters with the size of 1.6 micron and 1
23
micron were used respectively to filter the crude oil. The location of the unit is shown in
Figure 2.1. The crude oil was obtained from Ogallah Unit, Trego County, Kansas. The unit is
currently operated by Carmen Schmitt, Inc. The unit produces from Arbuckle formation
(3950-4060 ft). Reservoir temperature ranges from 92oF to 130oF with an average
pressure of 111 oF. Active water drives have maintained reservoir pressure at
approximately 1150 psi.
Figure 2-1 Ogallah Unit, Trego County, Kansas
A compositional analysis of the crude oil using Gas Chromatography (GC) technique was
performed by Core Labs and shown in Figure 2.2. % Asphaltenes (heptane insolubles) was
determined approximately 0.93 % based on ASTM D 893-85.
24
0.12
0.10
0.08
Mole fraction
0.06
0.04
0.02
0.00
C2
C6
C8
C36+
i-C4
i-C5
N2
H2S
C10
C12
C14
C16
C18
C20
C22
C24
C26
C28
C30
C32
C34
Figure 2-2 GC compositional analysis result of Ogallah crude oil
Core Labs also provided physical properties of the oil and the lumped heavy component
C36+, as presented in Table 2.1.
Molecular Weigh, g/mol 228.71

API 33.34
Density @ 14.7 psi & 60oF, g/cc 0.8584
Viscosity @ 14.7 psi & 60oF, cp 13.4
C36+ molecular weight, g/mol 873.24
C36+ density @ 14.7 psi & 60oF, g/cc 0.9978
Table 2.1 Physical properties of Ogallah crude oil and the lumped heavy component C36+
Commercial CO2 of 99.99 % purity was used
25
2.2 Slim Tube Displacements
The basis of slim tube tests is that the small-diameter tube filled with an unconsolidated
porous medium serves as an idealized medium for CO2 and crude oil to contact and develop
dynamic miscibility. Non-idealities such as viscous fingering and gravity effect are ignored
because of the large length to diameter ratio in slim tube configuration. Oil recovery is
therefore attributed to the thermodynamic phase behavior of the system. The recovery
performance at different pressures can be used to determine the MMP. Slim tube tests have
not only been used to determine reservoir candidates for miscible processes but also
widely used for fine tuning the reservoir simulator.
In this study, a number of slim tube displacements were conducted for a range of
pressures, holding the temperature constant at the reservoir temperature (110oF-125oF) to
determine the MMP of this system. An oil recovery factor of at least 90% at 1.2 HCPV of CO2
injected is used to define the MMP of the system.
26
2.2.1 Experimental Setup and Specifications
DPT PT
Test Oil
Densitometer
BPR
Gas
Oil
ISCO Pump ISCO Pump

Electronic Balance
Figure 2-3 Schematic of slim tube setup
Schematic of the slim tube setup is shown in Figure 2.3. Temperature of the system is
controlled and maintained in a Linberg/Blue M oven with Eurotherm temperature
controller. Pressure of the system is controlled and maintained by a back pressure
regulator at the outlet. Back pressure regulator models BPR-50 is a dome-load type, which
controls the upstream back pressure to whatever pressure is applied to the dome. It is
designed to operate using compressed gas in the dome and water, oil, gas in the body. A
27
high pressure bottle of inert gas, such as nitrogen, is required to pressurize the unit. The
back pressure regulator has a working pressure of 5000 psi at 200oF.
Three Valydine pressure transducers are installed to measure pressures at different
locations, such as pressure drop across the slim tube, upstream pressures (CO2/oil
pressure), and downstream pressure (back-pressure regulator pressure). The transducers
have the capability of measuring pressures up to 2500 psi with the accuracy of 0.25% of
their full scale (0-2500 psi).
The injection system consists of two Isco, Inc. 260DM syringe pumps (for CO2/crude oil
transfer and injection at a desired rate) and a transfer cylinder (for crude oil storage). The
capacity of the transfer cylinder is 485 cc. The cylinder can withstand a maximum pressure
of 3000 psi.
The slim tube consists of a coiled 38.29 ft-long stainless steel tube with an ID of 0.24 in.
packed with glass beads. Slim tube properties were evaluated by Rahmatabadi [12] and
listed in Table 2.2.
Length, ft 38.29
O.D, in 0.31
I.D, in 0.24
Porosity 0.37
Bulk volume, cc 347.8
Pore volume, cc 127.76
Permeability, mD 4900
Packing beads No. 2024
Table 2.2 Slim tube properties
28
Density of the effluent is measured continuously by an inline densitometer. The
densitometer consists of two units. The DPRn 422 density transducer measures the
characteristic frequency of vibration. The Anton Paar mPDS 2003V3 Evaluation unit
translates the characteristic frequency of vibration into a density value. The measuring
range is 0-3g/cc within the temperature range of -13oF – 257oF and the pressure range of
0-2900 psi.
Effluent is continuously flashed to atmospheric conditions. The separator gas is connected
to a flow meter. The separator liquid is collected in a graduated cylinder. The graduated
cylinder is placed on an electronic balance which is connected to the data acquisition
system.
2.2.2 Experimental Procedures
At least 2 PV of methylene chloride followed by 2 PV of mineral oil were injected to clean
the slim tube prior to the experiment. The slim tube was then saturated with at least 2 PV
of crude oil at the desired temperature. While the slim tube was being saturated with crude
oil, the system was pressurized gradually to the desired operating pressure using the back
pressure regulator. Upstream pressure changed accordingly to the pressure of the back
pressure regulator as it was set. To prevent pressure from one side of the diaphragm of the
back pressure regulator from becoming significantly higher than the pressure on the other
side and damage the diaphragm, it was necessary to pressurize the system slowly, for
29
example, 0-100 psig, switch off the gas supply valve and wait for the upstream pressure to
catch up with the downstream pressure. Once the desired pressure was reached, the
system was allowed to equilibrate under pressure. Pressure of the CO2 pump was set
slightly above the pressure of the back pressure regulator. Temperature of the pump was
set at temperature of the system. CO2 flow rate was set at a constant of 0.05cc/min. This
corresponds to a Darcy velocity of 8 ft/day.
A log file was created to record the following parameters: temperature of the system,
pressure of the system (back-pressure regulator pressure/downstream pressure),
upstream pressures (CO2/oil pressure), pressure drop across the slim tube, weight of the
separator liquid, and separator gas flow rate. The initial and final volumes of CO2 in the
pump were recorded manually.
The experiment was terminated when at least 1.2 HCPV of CO2 at the temperature and
pressure of the pump were injected. The system was depressurized by venting the dome
load gas slowly. Residual oil in the slim tube was removed by following the same cleaning
procedure as mentioned earlier.
The entire experiment was then repeated several times at different pressures holding other
variables constant.
30
2.2.3 Results and Discussions
Slim tube displacements were conducted at 110oF and 125oF representing the range of
temperatures reported from the field. The effluent was observed when it was flashed to the
atmosphere during the displacement. A gradual change of color of the flowing fluid from
the oil to clear gas was observed for pressures above MMP. Two-phase flow was observed
for pressures below MMP. Gas breakthrough was seen earlier from displacements at
pressures below MMP compared with displacements at pressures above MMP.
The recovery results of slim tube displacements at 110oF and 125oF are presented in Figure
2.4 and Figure 2.5.
100
80
% Oil recovery
60
P = 1509 psig
P = 1417 psig
40
P = 1375 psig
P = 1323 psig
20 P =1222 psig
P = 1126 psig
P =1022 psig
0
0.0 0.5 1.0 1.5 2.0
PV of CO2 injected
Figure 2-4 Results of displacement tests at 110oF
31
100
80
% Oil Recovery
60
40
P = 1714 psig
P = 1598 psig
20 P = 1509 psig
P = 1304 psig
P = 1096 psig
0
0.0 0.5 1.0 1.5 2.0
PV of CO2 injection
Figure 2-5 Results of displacement tests at 125oF
% oil recovery was calculated as follows:
Weight of oil produced

% oil re cov ery = × 100%
Pore Volume × Oil density
Pore volume (PV) of CO2 injected was calculated as follows:
time × CO2 flow rate

PV =
Pore Volume
32
Percentage of oil recovery at 1.2 PV of CO2 injected was plotted against slim tube average
pressure to determine the MMP of the system at 110oF and 125oF, as shown in Figure 2.6.
MMP of the system were estimated to be 1350 psig at 110oF and 1650 psig at 125oF. As
expected, MMP increases with increasing temperature. This phenomenon is closely related
to the dependence of CO2 density on temperature and pressure and is explained in detail in
the discussion of the next experiment.
Slim tube results indicated that miscibility was not achievable at the current reservoir
pressure of 1150 psig; however, the recovery efficiency was relatively high, 78% - 83% of
the original oil in slim tube for a temperate range of 110oF to 125oF and at 1150 psig.
100
% Oil Recovery @ 1.2 HCPV CO2
90
80
70
injected
60
50
40 T = 125dF
T = 110dF MMP MMP
30
900 1100 1300 1500 1700
Test Pressure, psia
Figure 2-6 Minimum miscibility pressure determinations at 110oF and 125oF
33
Since the MMP of this oil was determined to be 1350 psig at 110oF and the near miscible
range was selected from 0.8MMP to MMP in this study, the near miscible region for this
particular oil is defined from 1100 psig to 1350 psig at 110oF.
100
T = 110oF
% Oil Recovery @ 1.2 HCPV CO2
90
80
70
injected
60
50
40
0.8 MMP MMP
30
900 1000 1100 1200 1300 1400 1500 1600
Test Pressure, psia
Figure 2-7 Near miscible region for Ogallah crude oil at 110oF
Figure 2.8 and Figure 2.9 show the density profile of the effluent at 110oF and 125oF.
Density of the effluent was the same as density of crude oil at 110oF and at slim tube
average pressure prior to CO2 breakthrough. The abrupt change in density of the effluent
corresponded to CO2 breakthrough. Density of the effluent was much less than density of
the oil but higher than density of pure CO2 at pressures below MMP, which is the result of
two-phase flow with liquid dispersed in the CO2 rich phase through the densitometer. For
instance, densities of pure CO2 at 1022, 1126 and 1222 psig and 110oF are 0.19001,
34
0.23306 and 0.29009 g/cc, compared with densities of the effluent at the corresponding
pressures and temperature, 0.296, 0.415, and 0.527 g/cc.
0.9
0.8
0.7
Density [gm/cc]
0.6
0.5 P = 1509 psig

P = 1417 psig
0.4
P = 1375 psig
0.3 P = 1323 psig
0.2 P = 1222 psig
P = 1126 psig
0.1
P = 1022 psig
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
PV of CO2 injected
Figure 2-8 Density profile of the effluent at 110oF
35
1.0
0.9
0.8
Density (g/cc)
0.7
0.6
P = 1714 psig
0.5 P = 1598 psig
0.4 P = 1509 psig
P = 1304 psig
0.3
P = 1096 psig
0.2
0 0.2 0.4 0.6 0.8 1 1.2
PV of CO2 injection
Figure 2-9 Density profile of the effluent at 125oF
2.2.4 Conclusions
 MMP was estimated to be 1350 psig at 110oF and 1650 psig at 125oF. Miscibility is
therefore not achievable at the current reservoir pressure of 1150 psig.
 Near miscible range is defined from 1000 psig to 1350 psig at 110oF in this study.
 78% - 83% of the original oil in place was recovered at the current reservoir
pressure of 1150 psig and at a temperature range of 110oF to 125oF.
36
2.3 Swelling/Extraction Tests
Swelling/extraction tests were performed to examine the oil recovery mechanisms in the
near-miscible region and to provide data to tune the phase behavior model. Swelling tests
were conducted to determine the relationship between saturation pressure, swelling factor
and CO2 volume injected. Extraction tests were carried out to examine the extraction of
liquid hydrocarbon into a CO2-rich phase and the effect of pressure on the extraction.
2.3.1 Experimental Setup and specifications
Figure 2-10 Experimental setup include (1) Gas cylinder (2) Isco Syringe pump (3) Fisher Isotemp
circulator (4) Fisher Isotemp Immersion circulator (5) water bath (6) high pressure view cell (7)
mixing bar (8) laboratory jack (9) computer (10) cathetometer with telescope (11) vacuum pump
[13]
37
The schematic of swelling/extraction setup is shown in Figure 2.10. An Isco, Inc. 100DM
syringe pumps is used for CO2 injection. Temperature of the pump is controlled by a Fisher,
Inc. Isotemp circulator, model 3016 and measured by an Ertco-Eutechnic 5 digit
thermister, model 4400 in the range of 0-100oC.
The gas lines are heated using fiberglass covered heating tape, controlled by two variable
AC transformers, Staco Energy model 3PN1010B. Temperature of the gas lines is measured
using T-type thermocouples. Fiberglass cloth tape is used to prevent heat dissipation to the
surroundings.
The key component of this setup is the high pressure view cell with high pressure gauge
glass window allowing visual observations of fluids under experimental conditions. The
view cell is made of stainless steel and has a volume of 26 cc. The gauge glass window
allows a maximum temperature of 280oC and pressure of 4000 psi. Pressure in the view
cell is measured by a 5000 psi Heise DXD Series 3711 precision digital pressure transducer.
A 3.2mm diameter × 12.7 mm PTFE coated stir bar is placed inside the view cell. Mixing is
achieved by an external rare-earth magnet in a slot behind the cell raised and lowered by a
pulley system.
The view cell is immersed into the water bath by raising/lowering the platform jack. The
temperature of the water bath is adjusted by an immersion circulator Haake DC30/DL3
and a Fisher, Inc. Isotemp circulator, model 3016.
38
An Eberbach 5160 cathetometer is used to measure the height of the liquid in the view cell.
Figure 2-11 An actual image of the swelling/extraction experimental setup
The pump was filled with CO2. Temperature of the pump was set constant and above the
critical temperature of CO2 (31.1oC). Pressure of the pump was set constant at the
maximum anticipated pressure. The pump automatically adjusted the volume of CO2 to
achieve constant temperature and pressure. Temperature of the gas lines was set at
39
temperature above the critical temperature of CO2 to avoid CO2 condensation inside the
lines. Temperature of the water bath was set constant at the desired temperature.
A predetermined volume of crude oil was carefully injected into the view cell to avoid
liquid droplets on the wall of the view cell. The view cell was attached to the gas lines and
then immersed into the water bath.
When the whole system was thermally equilibrated, the gas lines and the view cell were
quickly flushed with CO2 at low pressure to remove any residual gas or air.
A log file was created to record the following parameters: the pump condition
(temperature, pressure and volume of CO2), the temperature of the gas lines, the view cell
condition (temperature and pressure). The height of the liquid sample in the view cell was
recorded manually. Initially conditions should also be recorded manually.
The cell pressure was increased in discrete steps by CO2 injection from the top of the view
cell. CO2 injection was stopped when a desired pressure was achieved. CO2 flow rate was
kept slowly so that there was no PVT disturbance. This was done by checking the pump
pressure frequently to make sure the pump pressure did not drop too much from the set
pressure. Final volume of CO2 in the pump was recorded when CO2 flow rate is read zero.
During pressurization process, the time required for the contents in the view cell to
equilibrate under a particular pressure and temperature is minimized by magnetically
stirring. At that time, the following parameters were recorded manually: the height of the
40
liquid sample in the view cell, the pump condition (temperature, pressure & final volume of
CO2), temperature of gas lines and the view cell condition (temperature & pressure).
In the end, the view cell was cleaned with methylene chloride, acetone solution and blown
dry with compressed air.
2.3.3 Experimental Principles
The phase equilibria data were obtained based on a mass balance and the following
assumptions:
 The pressure of CO2 is much greater than the vapor pressure of the crude oil
 Vapor phase composition of the hydrocarbon component is much less than CO2
The mass balance equation of CO2 is as follows:
m g = m pump − mlines − mheadspace + m o lines + m o headspace

m pump = ∆V pump ρ (T pump , Ppump )
mlines = ∆Vlines ρ (Tlines , P)
mheadspace = ∆Vheadspace ρ (T , P) = (Vcell − Vliquid )ρ (T , P)
m o lines = ∆Vlines ρ (Tlines , P o )
m o headspace = ∆Vheadspace ρ (T , P o ) = (Vcell − Vliquid )ρ (T , P 0 )
Where
mg is the mass of CO2 dissolved in the liquid
41
mpump is equal to the product of volume of CO2 displaced from the pump and density of CO2
at the pump constant temperature & pressure
mlines is equal to the product of volume of the lines and density of CO2 at temperature of the
lines & system equilibrium pressure
molines is equal to the product of volume of the lines and density of CO2 at temperature of the
lines & system initial pressure
mheadspace is equal to the product of volume of the headspace and density of CO2 at
temperature & pressure of the equilibrium system. The volume of the headspace is the
difference between volume of the cell and volume of the liquid in the cell.
moheadspace is equal to the product of volume of the headspace and density of CO2 at
temperature & pressure of the initial system. The volume of the headspace initially is the
difference between volume of the cell and initial volume of the liquid in the cell.
CO2 density was calculated using REFROP database which used the ultra-accurate Span-
Wagner equation of state.
Mole fraction of CO2 in the liquid phase was calculated as follows:
ng mg / M g
xg = =
n g + nl m g / M g + ml / M l
42
2.3.4 Apparatus Validation
The apparatus was verified by comparing experimental data obtained from this apparatus
with literature data for n-decane/CO2 mixture at 71.1oC by Ren et al. [13]. The
experimental data had excellent agreement with the literature data obtained from different
experimental methods.
Phase equilibrium data of CO2/n-decane at 71.1oC was obtained and shown in Table 2.3.
Run 1 Run 2
Pressure, psi xCO2 Pressure, psi xCO2
192.18 0.09609 209.43 0.10900
425.25 0.21070 434.97 0.22498
701.98 0.34940 643.68 0.32669
1005.55 0.50260 866.89 0.44391
1270.24 0.61657 1080.39 0.52821
1448.35 0.69121 1321.58 0.64927
1531.45 0.72330 1495.77 0.71555
1666.05 0.77723 1612.38 0.76148
1709.41 0.80048
1771.35 0.82829
Table 2.3 Phase equilibrium data of CO2/n-decane at 71.1oC
Analysis of this data was based on the assumption that the amount of liquid component in
the vapor phase is negligible. Although the composition of the vapor phase was not actually
analyzed in our experiments, it had been demonstrated earlier by Ren et al. [13] that the
percentage of n-decane in CO2 vapor phase was less than 0.13%. Figure 2.12 shows that
the p-x phase equilibrium of CO2/n-decane generated using this apparatus were
reproducible and in excellent agreement with literature data, Nagarajan & Robinson, Jr.
[14] and Jennings & Schucker [15], and therefore, it could be used for crude oil/CO2 system.
43
2000
This work #1
This work # 2
1600 Nagarajan & Robinson Jr.
Jennings & Schucker
Pressure (psi)
1200
800
400
0
0.0 0.2 0.4 0.6 0.8 1.0
Mole fraction of CO2 in the liquid phase
Figure 2-12 Comparison of liquid phase compositions for CO2+ n-decane system at 71.10oC (160oF)
with literature data (○) this work (□) this work (∆) Nagarajan & Robinson Jr. (◊) Jennings & Schucker
Effect of each variable, temperature, pressure and initial volume on CO2/oil phase behavior
was investigated while holding others constant.
2.3.5.1 Effect of System Pressure
44
Figure 2-13 Change of initial oil volume with pressure
Figure 2.13 illustrates the observations of Ogallah oil/CO2 phase behavior. Volume of the
fluid in the view cell increased with increasing pressure since CO2 dissolved in and swelled
the oil. As pressure was further increased, CO2 density increased. Since dense-phase CO2
had the ability to extract hydrocarbon components from crude oil more easily than if it
were in the gaseous phase, CO2 started vaporizing or extracting hydrocarbons from crude
oil, volume of the fluid in the view cell was therefore reduced.
45
1.4 0.8
Swelling Factor
1.2 0.7
CO2 solubility
0.6
1.0
Swelling Factor
CO2 solubility
0.5
0.8
0.4
0.6
0.3
0.4
0.2
0.2 0.1
0.0 0.0
0 500 1000 1500 2000 2500
Pressure, psi
Figure 2-14 Effect of pressure on CO2 solubility and swelling factor at 110oF
Figure 2.14 shows the swelling/extraction curve for Ogallah/CO2 system at 110oF with the
sample size of 3 cc. The sample volume was about 11 % volume of the view cell. The
swelling factor (SF) of oil is the ratio of liquid volume at test pressure divided by the liquid
volume at atmospheric pressure and at 110oF. This value is determined by measuring the
change of the interface level as a result of CO2 dissolution in the oil or as a result of
hydrocarbon extracted into the CO2 rich vapor phase. Swelling factor was equal to 1 at
initial conditions. As a result of CO2 dissolution into the liquid phase, the liquid phase
swelled and the swelling factor was greater than 1. Maximum swelling occurred at 1158
psi, when volume of the liquid phase became 1.21 of its original volume with 0.728 mole
46
fraction of CO2 dissolved in the liquid phase. Major extraction started at approximately
1158 psi. As pressure increased, hydrocarbon components of the crude oil were removed
from the liquid phase, the liquid phase shrank and swelling factor was reduced. At 2035.2
psi, the volume of CO2 rich liquid phase shrank as much as 39.2 % of its original volume.
CO2 solubility is also plotted in Figure 2.15 as a function of pressure up to 1158 psi.
Calculations of CO2 solubility at pressures above 1158 psi are invalid since the assumption
that the components of the liquid phase do not vaporize does not hold true.
Extraction or vaporization of hydrocarbon components from crude oil appears to be the
primary oil recovery mechanism in the near miscible range, 1100 psig to 1350 psig.
2.3.5.2 Effect of System Temperature
Swelling/extraction experiments were performed under various temperatures from 105oF
to 125oF.
47
1400
1200
1000
Pressure, psia
800
600
105F
400 110F
115F
200 120F
125F
0
0.0 0.2 0.4 0.6 0.8 1.0
Figure 2-15 Effect of temperature on CO2 solubility
1.4
1.2
1.0
Swelling Factor
0.8
0.6
105F
0.4 110F
115F
0.2 120F
125F
0.0
0 500 1000 1500 2000 2500
Pressure, psi
Figure 2-16 Effect of temperature on Swelling/ Extraction curves
48
Effects of temperature on CO2 solubility and oil swelling/extraction curve are shown in
Figure 2.15 and Figure 2.16. CO2 solubility increased with increasing pressure and
decreased with increasing temperature. The rate of oil swelling decreased with increasing
temperature. The pressure at which oil swelling reached maximum or at which CO2 began
extracting components from crude oil increased with increasing temperature, ranging from
1158.8 psi to 1259.6 psi at a temperature range of 105oF to 125oF. The rate of oil shrinkage
decreased with increasing temperature.
The temperature dependence of CO2/Ogallah crude oil phase behavior is explained by the
dependence of CO2 density on pressure and temperature. Density of CO2 as a function of
pressures and temperature are calculated using REFPROP and presented in Figure 2.18. It
takes higher pressure to achieve the equivalent CO2 density at a higher temperature; and
therefore, the pressure at which extraction starts increases with increasing temperature.
The rate of change of CO2 density with pressure is faster at lower temperature which
explains why the extraction rate is faster at lower temperature. Mass transfer between CO2
and oil phase is necessary to achieve dynamic miscibility. Since mass transfer between CO2
and oil phase starts earlier and faster at a lower temperature, MMP decreases with
decreasing temperature.
49
0.8
T = 110oF
0.7
T = 125oF
0.6
Density, g/cc
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000
Pressure, psi
Figure 2-17 Dependence of CO2 density on temperature and pressure (REFPROP)
2.3.5.3 Effect of Initial Oil Volume
Swelling/Extraction tests were performed with different initial oil volume, such as 3cc
(11% view cell volume), 9cc (35% view cell volume) & 14cc (54% view cell volume) at
110oF to investigate its effect on CO2 solubility and swelling factor.
50
1400
1200
1000
Pressure, psi
800
600
400
14cc
200 9 cc
3 cc
0
0.0 0.2 0.4 0.6 0.8
Figure 2-18 Effect of initial oil volume on CO2 solubility at 110oF
1.6
1.4
1.2
Swelling factor
1.0
0.8
0.6
0.4 14 cc
0.2 9 cc
3 cc
0.0
0 500 1000 1500 2000 2500
Pressure, psi
Figure 2-19 Effect of initial oil volume on swelling/extraction curve at 110oF
51
The effect of initial oil volume on CO2 solubility and swelling factor at 110oF are shown in
Figure 2.18 and Figure 2.19. When the initial volume of crude oil in the view cell was large,
the amount of hydrocarbons extracted from the oil phase is considerably lower than when
the initial volume of crude oil in the view cell is small. The pressure at which significant
extraction began decreased with decreasing initial cell oil volume.
Similar observations have been reported by Hand et al. [16] and were first noted by Holm
and Josendal [4] in their work where a 30% of initial volume of sample was used for the
swelling test. Hand et al. [16] explained the effect of initial oil volume on
swelling/extraction curve by plotting the change in cell global gas composition with
pressure on a p-x diagram for the Flounder oil/CO2 system at 233oF.
52
Figure 2-20 Computed p-x diagram for the Flounder oil/CO2 system at 233oF [16]
The compositional paths for swelling/extraction tests with initial oil volumes of 40 and
60% display only oil swelling because the compositional paths cross only increasing-liquid-
volume-quality lines with increasing system pressures until they terminate in a bubble
point. The 10 and 20% initial oil volume compositional paths that cross increasing-liquid-
volume-quality lines at lower cell pressures are shown to cross decreasing-liquid-volume-
quality lines at higher pressures. These tests display oil swelling first and then oil
shrinkage with increasing pressures.
53
1400
1200
1000
Pressure (psia)
800
600
400
200
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mole fraction Carbon Dioxide
2-Phase boundary 80% Oil 40% Oil

30% Oil 20% Oil 10% Oil
3 cc (11% Oil) 9 cc (35% Oil) 14 cc (54% Oil)
Figure 2-21 Computed p-x diagram for Ogallah oil/ CO2 system at 110oF
Figure 2.21 shows a predicted p-x diagram for Ogallah oil/CO2 system at 110oF by
Winprop, CMG software. Also plotted are the experimental results, pressure vs. mole
fraction of CO2 for systems of 11%, 35% and 54% oil / CO2 at 110oF. Mole fraction of CO2
was determined by total mole of CO2 injected into the system divided by sum of total mole
of CO2 injected into the system and the moles of crude oil. The difference of
swelling/extraction curves between sample sizes is explained likewise, because the
compositional paths cross the liquid-volume-quality lines differently with increasing
pressure.
54
2.3.6 Conclusions
 CO2 solubility increases with increasing pressure and decreases with increasing
temperature. As a result of CO2 dissolution into the liquid phase, the liquid phase
swells. The degree of swelling depends on pressure and temperature.
 The pressure at which CO2 begins to extract significant amounts of hydrocarbons
from crude oil increases with increasing temperature and with increasing initial oil
volume in the view cell. The amount of extraction increases with increasing vapor-
phase volume (decreasing initial oil volume) and decreases with increasing
temperature.
 Significant extraction started at pressure of 1150 psig & 110oF. Extraction or
vaporization of hydrocarbons into the CO2 rich phase is the primary mechanism for
oil recovery in the near-miscible region, from 1100 psig to 1350 psig & at 110oF
 Swelling/extraction tests provide valuable phase behavior data which is used later
to tune the phase behavior model.
2.4 Viscosity Measurements
A Cambridge Applied Systems’ high pressure viscometer (ViscoPro 2000 System 4-SPL-440
with Viscolab software) utilizes an electromagnetic concept to measure viscosity of pure
crude oil and/or crude oil/CO2 mixture. The accuracy of this viscometer was validated
earlier by Ahosseini et al. [17], first by using standard calibration solutions, then by
55
comparing viscosity values of n-hexane measured using this viscometer with literature
values.
2.4.1 Principle of Operation
The viscometer utilizes the principles of annular flow around an axially oscillating piston.
It contains two magnetic coils inside a stainless steel body. A low mass stainless steel
piston inside the measurement chamber is magnetically forced back and forth in the fluid.
As the piston is pulled toward the bottom of the measurement chamber, it forces the fluid
at the bottom of the chamber to flow around the piston toward the sensor opening where it
interchanges with the normal flow of the fluid. On the upward piston stroke, fresh process
fluid is pulled around the piston to the bottom of the measurement chamber. The time
required for the piston to move a fixed distance is related to the viscosity of the fluid in the
chamber. Temperature is measured continuously with the use of a platinum Resistance
Temperature Detector (RTD) mounted at the base of the measurement chamber.
56
2.4.2 Experimental Setup and Specifications
P
CO2 Inlet R
Liquid Outlet
Liquid Inlet
High pressure sensor
R
T
Circulation
Pump
Oven
Liquid Outlet
High Pressure Generator
Figure 2-22 Schematic of high-pressure viscosity measurement setup
The schematic of the setup is shown in Figure 2.22. The high pressure sensor is placed
inside a temperature-controlled oven and connected to a manual high pressure syringe
pump (Model No. 50-575-30; 30,000 psi). The sensor is capable of measurements from 0.2
to 10,000 cp at a maximum pressure of 137.9 MPa and in a temperature range of 233.15K
to 463.15K .The measurement chamber of the high pressure sensor is connected to a
rupture disk (RD) and a precision pressure transducer (PT). When the setup is used for
viscosity measurements of pure crude oil as a function of pressure, a high-pressure
generator is required. When it is used for viscosity measurements of crude oil/CO2 mixture,
the pressure of the system is increased by CO2 injection. ISCO 260D pump- Model 1020
BBB-4 is used for CO2 transfer/injection. A view cell placed inside the oven allows
57
observations of crude oil/CO2 interaction. During pressurization process, the time required
for the contents in the system to equilibrate under a particular pressure and temperature is
minimized by a circulation pump-Micropump, Inc. Model 415A.
Figure 2-23 An actual image of the high-pressure viscosity measurement setup
58
Experimental procedures to measure oil viscosity and oil/CO2 mixture are outlined in the
following sections.
2.4.3.1 Oil Viscosity Measurement Procedure
Figure 2-24 A high pressure generator
The schematic in Figure 2.24 illustrates a very basic layout for using a high pressure
generator. A reservoir (R) is shown connected by means of valves and fittings to a
component (C) that is to be pressurized. With valve “B” closed and valve “A” open, the
handle of the high pressure generator is rotated counter-clockwise to create enough space
inside the cylinder body to store the crude oil. Valve “A” is then closed and valve “B” is
opened. By rotating the high pressure generator handle clockwise slowly, the piston will
compress the fluid to develop pressure in the component that is to be pressurized. If
sufficient pressure is not reached in one stroke, the system can be “recycled.” Valve “B” can
59
be closed in order to maintain pressure in the components. Valve “A” is then opened, and
fluid is again drawn into the Pressure Generator from the reservoir. Closing Valve “A” and
opening Valve “B” will now allow the Pressure Generator to be operated to develop
increased pressure in the component.
In this experimental setup, the oil was pumped into the cylinder body using the manual
high pressure syringe pump until the cylinder body is full. Excessive fluid introduction
would cause pressure buildup in the cylinder body; therefore, the system outlet must have
been opened before Valve “B” was opened to avoid damaging the sensor. The oil was
introduced into the system and into the measurement chamber, opening both Valve “A” and
Valve “B” and bypassing the high pressure generator. Temperature of the oven was set at
the desired temperature. When the system was at thermal equilibrium, the system was
purged to dislodge any potential bubbles. Temperature reading was taken from the RTD of
the viscometer. Valve “A” was closed, Valve “B” was opened when the system was
pressurized. The viscometer also needed to be turned off. By rotating the handle of the high
pressure generator slowly clockwise, pressure was developed in the system until a desired
pressure had reached. Viscosity reading was taken when temperature and pressure of the
system is stable.
60
2.4.3.2 Oil/CO2 Mixture Viscosity Measurement Procedure
Temperature of the oven is set constant at the desired temperature. Temperature of ISCO
pump was set constant at the desired temperature. Pressure of the pump was set constant
at the maximum anticipated pressure. Liquid was introduced into the measurement
chamber and into the view cell. When the system was at thermal equilibrium, the system
was purged to dislodge any potential bubbles. Temperature reading was taken from the
RTD of the viscometer. Viscometer needed to be turned off when CO2 was introduced into
the system. The system pressure was increased in discrete steps by CO2 injection from the
top of the view cell. CO2 injection was stopped when a desired pressure was achieved.
During pressurization process, a micropump was used to circulate the liquid inside the
system to help accelerating the mass transfer of the gas phase into the liquid phase.
Reading was taken when temperature and pressure of the system were stabilized.
61
6.0
5.6
Viscosity (cp)
5.2
4.8
4.4
4.0
0 500 1000 1500 2000
Pressure (psi)
Figure 2-25 Effect of pressure on Ogallah viscosity at 110oF
5.0 0.8
Viscosity Run1
0.7
4.0 Viscosity Run2
Viscosity Run3 0.6
Viscosity (cp)
CO2 solubility
3.0
0.5
CO2 Solubility
0.4
2.0
0.3
0.2
1.0
0.1
0.0 0
0 500 1000 1500 2000
Pressure (psi)
Figure 2-26 Effect of CO2 dissolution in crude oil on the viscosity of Ogallah fluid
62
Figure 2.25 shows the relationship of viscosity of pure crude oil with pressure at 110oF.
Figure 2.26 demonstrates the effect of CO2 dissolution into crude oil on viscosity of crude
oil at 110oF. As a result of CO2 dissolution into the liquid phase, viscosity of the liquid phase
was reduced. At pressures above 1000 psig, the effect of CO2 on oil viscosity is minimal. The
viscosity of the liquid phase was reduced as much as a factor of five in the near miscible
region, from 1100 psig to 1350 psig.
2.4.5 Conclusions
 A nearly five-fold reduction in oil viscosity was achieved in the near-miscible region
with CO2 injection.
 The reduction of oil viscosity improved the total mobility ratio between reservoir oil
and the displacing fluid, which will affect the recovery efficiency in a favorable way.
 Viscosity measurements are useful in tuning the phase behavior model.
2.5 Phase Behavior Model
Equations of State (EOS) have been introduced widely to model and predict CO2-crude oil
phase behavior. An EOS is an algebraic relationship between pressure, temperature, and
molar volume for a single component or a mixture. Peng-Robinson EOS was used in this
study to describe the fluid properties and the oil/CO2 interaction.
63
RT a
p= − 2
v − b v + 2vb − b 2
The parameters a and b are defined using the following mixing rule for mixtures,
a = ∑ xi Si
i
Si = ai ∑ x j (1 − d ij ) a j
j
b = ∑ xibi
i
Where, dij is an empirically determined interaction coefficient.
For pure components, the parameter ai and bi are expressed in terms of the critical
properties and the accentric factor:
a = acα
ac = Ω a (RTc ) / pc
(
α = 1 + κ 1 − T / Tc )
b = Ωb RTc / pc
Where, Ωa and Ωb are the EOS parameters with the default values of 0.45723553 and
0.077796074 for the PR EOS.
The κ is obtained from the following empirical correlations,
κ = 0.379642 + 1.48503ω − 0.164423ω 2 + 0.016666ω 3
While the application of the EOS to simple mixtures is relatively straightforward, crude oil
systems pose many seemingly insurmountable problems. The essential infinite number of
64
components contained in a typical crude oil makes it impossible to obtain a complete
chemical analysis. Standard oil analysis lumps all components heavier than C36 into a
single C36+ pseudo-component characterized by its average molecular weight and density.
Therefore, the parameters of the EOS were adjusted so that the EOS could reproduce the
laboratory-determined oil properties and the observed oil/CO2 interaction.
Slim tube model was constructed using 1-D compositional simulator with the tuned EOS.
Simulated slim tube results were compared with experimental slim tube results. The
following sections describe in detail how to develop the phase behavior model/slim tube
model in CMGTM software and how to characterize the phase behavior model to match with
experimental PVT data and slim tube results.
2.5.1 Phase Behavior Modeling using WINPROP
WinProp, a CMGTM software was utilized to build the phase behavior model. The EOS
requires the following properties for each component, critical pressure (Pc), critical
temperature (Tc), acentric factor (ω), and interaction coefficients between different
components (dij) to perform phase equilibrium calculations. The molecular weight (MW) is
also required to calculate mass densities. Additional factors such as the volume shifts τ, and
the equation-of-state parameters Ωa and Ωb can also be adjusted for each component to
enhance the equation of state predictions.
65
Pure hydrocarbon components were selected from the software library, as well as
generalized single carbon number (SCN) petroleum fractions FC6 through FC45. The
specific gravities, molecular weights and boiling points of the SCN fractions were taken
from Whitson (1983). The critical properties of these fractions were calculated with the
Lee-Kesler correlation. The heavy lumped component C36+ was defined using specific
gravity and molecular weight. Physical and critical properties of C36+ were assigned using
Twu and Lee-Kesler correlation respectively. For accentric factors, the Lee-Kesler
correlation was recommended for petroleum fractions. Equilibrium phase viscosities were
calculated with the Pedersen viscosity corresponding states model. The Pedersen viscosity
correlation uses the principle of corresponding states to calculate the viscosity of a
component or mixture, knowing the viscosity of a reference substance at the same
conditions of reduced pressure and temperature. The deviation from simple corresponding
states is accounted for by a “rotational coupling coefficient”, α. The reference substance for
the Pedersen model is methane. The viscosity of the mixture is calculated according to the
following formula:
𝜇𝜇𝑚𝑚𝑚𝑚𝑚𝑚 (𝑃𝑃, 𝑇𝑇) 𝑇𝑇𝑐𝑐,𝑚𝑚𝑚𝑚𝑚𝑚 −1/6 𝑃𝑃𝑐𝑐,𝑚𝑚𝑚𝑚𝑚𝑚 2/3 𝑀𝑀𝑀𝑀𝑚𝑚𝑚𝑚𝑚𝑚 1/2 𝛼𝛼𝑚𝑚𝑚𝑚𝑚𝑚
=
𝜇𝜇𝑜𝑜 (𝑃𝑃𝑜𝑜 , 𝑇𝑇𝑜𝑜 ) 𝑇𝑇𝑐𝑐,𝑜𝑜 𝑃𝑃𝑐𝑐,𝑜𝑜 𝑀𝑀𝑀𝑀𝑜𝑜 𝛼𝛼𝑜𝑜
When the components representing the fluid model had been selected and their
compositions had been specified, a grouping scheme was performed primarily for the
purpose of speeding up the simulation running time. Whitson suggested that C7+ should
be grouped into NH pseudo-components given by,
66
N H = 1 + 3.3 log( N − 7 )
The groups are separated by molecular weights MI given by,
(
M I = M C7 M N / M C7 )
I / NH
where N = CN of the heaviest fraction in the fluid description
I = 1 to NH
Therefore,
NH = 5 pseudo-groups
Group 1 < M1 = 138.303
M1 <Group 2 < M2 = 191.276
M2 < Group 3 < M3 = 264.539
M3 < Group 4 < M4 = 365.865
Finally a 8-component EOS fluid model was obtained after grouping:
C3 + iC4 + nC4 (1)
nC5+ iC5 (2)
C6 (3)
C7 – C9 (4)
C10 – C13 (5)
C14 – C18 (6)
C19 – C25 (7)
C26 – C36+ (8)
67
2.5.2 Equation of State Characterization
Figure 2.28 and Figure 2.29 present the experimental data and the simulated data after
tuning EOS parameters. Molecular weight (MW) of the heavy fraction was adjusted to
match the oil density. Coefficients of Pedersen viscosity correlation were adjusted to match
the oil viscosity. Binary interaction coefficients (BIC) between CO2 and the hydrocarbon
components as well as CO2 volume shift factor were adjusted to match saturation pressure
and swelling data. Table 2.5 shows the adjustment of each parameter to achieve the match.
Lower Upper Initial Final

Variable Bound Bound Value Value
MW C26 to C36+ 5.07E+02 7.60E+02 6.34E+02 5.98E+02
Volume shift of CO2 -1.54E-01 9.20E-01 0.00E+00 9.11E-01
BIC (C7-C9)-CO2 0.00E+00 2.00E-01 1.50E-01 0.00E+00
BIC (C10-C13)- CO2 0.00E+00 2.00E-01 1.50E-01 0.00E+00
Coefficient of viscosity correlation
#4 1.00E+00 2.22E+00 1.85E+00 1.04E+00
Coefficient of viscosity correlation
#5 4.14E-01 6.21E-01 5.17E-01 6.21E-01
Table 2.4 Adjustments of EOS parameters
68
53.0 6.0
5.0
Oil density (lbm/ft3)
52.5
Oil viscosity (cp)

4.0
52.0 3.0
Density_Model 2.0
51.5 Density
Viscosity_Model 1.0
Viscosity
51.0 0.0
0 500 1000 1500 2000
Pressure (psia)
Figure 2-27 Comparison of viscosity/density experimental data and simulated data before/after EOS tuning
69
1400 1.40
Psat
Saturation Pressure (psia)
1200 1.35
Exp. Psat
S.F. 1.30
1000
Swelling Factor
Exp. S.F. 1.25
800
1.20
600
1.15
400
1.10
200 1.05
0 1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
Figure 2-28 Comparison of saturation pressure/swelling factor experimental data and simulated data
before/after EOS tuning
The maximum percentage error between simulated data and experimental data after
tuning the EOS is presented in Table 2.6
Maximum error percentage

Viscosity 3%
Density 1%
Saturation pressure 7%
Swelling factor 1%
Table 2.5 Maximum percentage error between simulated and experimental data
2.5.3 Slim Tube Modeling using GEM
A slim tube model was created in compositional model simulator of GEM from CMG
package and the phase behavior model of the reservoir was imported in it as the original
70
saturated fluid of the tube. The slim tube was represented by a one dimensional linear
model using 320 grid blocks. The grid block sizes were 0.125 ft, 0.0185 ft and 0.0185 ft in I,
J, K direction respectively. One injector and one producer were incorporated at the ends of
the model.
Length, ft 40
Porosity, PV 0.367
Permeability, mD 4900
Pore Volume, cc 142.3
No. of blocks 320
Length of blocks, ft 0.125
Cross section, ft2 0.0003423
Each edge, ft 0.0185
Geometry Square Cross-section
Table 2.6 Slim tube model properties
The slim tube gas/oil relative permeability curves used in this model were obtained from
Negahban and Kremesec [18] and illustrated in Figure 2.27.
71
1.0
0.8
kr- Relative Permeability
0.6
0.4
0.2 krg vs Sg
kro vs Sg
0.0
0.0 0.2 0.4 0.6 0.8 1.0
Sg
Figure 2-29 Relative permeability curve of oil-gas used in the simulation [18]
A number of slim tube displacements were simulated and compared with experimental
slim tube results in Figure 2.30.
72
% Oil Recovery @ 1.2 HCPV CO2 100
90
80
70
injected
60
50
40 Experimental
MMP Simulated
30
800 1000 1200 1400 1600
Test Pressure, psi
Figure 2-30 Comparison of simulated MMP and experimental MMP
Lower oil recoveries at pressures below MMP as seen in Figure 2.30 indicated early gas
breakthrough. In order to match simulated slim tube results with experimental slim tube
results, the slim tube gas/oil relative permeability curves must be adjusted by keeping the
relative permeability end points and changing the curvature of the relative permeability
curves. Figure 2.31 shows the slim tube gas/oil relative permeability curves used in the
simulation.
73
1.0
0.8
kr- Relative Permeability
0.6
0.4
0.2
0.0
0.0 0.2 0.4 0.6 0.8 1.0
Sg
krg vs Sg [18] kro vs Sg [18]
krg vs Sg (Simulation) kro vs Sg (Simulation)
Figure 2-31 Relative permeability curve of oil-gas before and after adjusted to match slim tube results
Figure 2.32 compares the recovery efficiency from simulation and experimental work at 1.2
PV of CO2 injected. As shown by Figure 2.32, the phase behavior model predicts the MMP
and the oil recovery reasonably well.
74
1.0
0.9
Experiment
Oil Recovery (HCPV)
0.8 MODEL
0.7
0.6
0.5
0.4
1000 1100 1200 1300 1400 1500 1600
Pressure (psig)
Figure 2-32 Comparison of simulated MMP and experimental MMP after tuning the relative permeability
curve
75
3 Core Flow Tests
Core flow tests were conducted under conditions that are more realistic than in slim tubes
and closer to the field conditions. They are virtually identical in procedure to the slim tube
displacement tests, except that a core sample replaced the bead-packed tube. The method
of conducting the tests and the results monitored are identical to the slim tube test.
CO2 core floods are difficult to interpret since linear displacements may or may not exhibit
characteristics common to a given reservoir such as dispersion, mobile water, wettability,
viscous fingering, gravity segregation, oil bypassing due to heterogeneity, and trapped oil
saturation. However, Orr et al. [19] suggested that core flow tests do seem appropriate for
answering at least three questions directly (1) Can CO2 mobilize tertiary oil under
conditions that are close to field displacement conditions than those occurring in slim tube
displacements? (2) What is the residual oil saturation in the swept zone of a CO2
displacement? (3) Does CO2 injection alter reservoir permeability by asphaltene
precipitation or dissolution of minerals?
3.1 Cores
Cores from Arbuckle reservoirs are limited. Berea sandstone, Baker dolomite and Arbuckle
dolomite rock samples were used in the core flow tests. Berea sandstone and Baker
dolomite were quarried rock samples whereas Arbuckle dolomite was cored sample from
76
Hadley well, Bemis-Shutts Field at Ellis County, Kansas. Before cores were epoxy encased
and cast inside an aluminum cylinder with high strength epoxy, at least three
measurements of the diameter and the length of each core were obtained. The average
value of the diameter and the length was used to calculate the area and the bulk volume of
each core. The dry weight of the core holder was recorded.
Upon completion of each core flow test, the cores were cleaned to restore to its initial
condition and reused for the next experiment. The core cleaning procedure is outlined as
follows:
 The cores were successively cleaned with methylene chloride and methanol. At least
10 PV of methylene chloride followed by 10 PV of methanol was injected. The
sequence was repeated at least three times until the effluent was clear.
 The cores were blown dry by CO2 and vacuumed to remove residual air/CO2.
 The cores were fully saturated with either brine or crude oil, depending on which
type of core flow tests would be conducted. Brine was used for tertiary CO2 floods.
Crude oil was used for secondary CO2 floods.
77
Figure 3.1 shows the experimental setup for vacuuming and saturating the cores,
Figure 3-1. Vacuum and saturation setup
 VACUUM (700mmHg): Open Valve 1, Close Valve 2
 DRY WEIGHT MEASUREMENT: Close Valve 1 & 2, Disconnect the core holder with
the vacuum line, Weigh the core holder with all the fittings (Weight of all the fitting
would be subtracted later)
 The purpose of the syringe was to ensure that the line which immersed in the
graduated cylinder was clear of air.
 SATURATION: Close Valve 1, Open Valve 2 (Brine will be drawn in the core holder),
Wait until the level of the brine did not change (During the saturation process, the
line was always ensured to be immersed in the brine)
 PORE VOLUME ESTIMATION: Estimate the change of volume of brine in the
graduated cylinder before and after brine was suck into the evacuated cores (dead
volume from fittings must be subtracted).
78
3.2 Fluids
Brine with 1wt% total dissolved solids (TDS) was used. Composition of brine was 0.5 wt%
MgCl2 and 0.5 wt% CaCl2 in deionized water. Brine was filtered using 0.22 micron
membrane (Micron Separations Inc.) before use. Density of brine was obtained using the
densitometer. Viscosity of brine was measured using Brookfield DV-III Ultra programmable
rheometer. Density and viscosity of brine are summarized in Table 3.1.
T = 86oF T = 110oF
Brine density, g/cc 1.0007 0.99591
Brine viscosity, cp 0.8350 0.7250
Table 3.1 Density/Viscosity of Brine at 86oF/110oF
3.3 Equipments and Procedures
Experimental setup and procedure for core characterization and core flow tests are
described in the following sections.
3.3.1 Core Characterization
Core characterization includes determining pore volume, porosity and permeability for
each core. The following sections discuss in detail how to obtain those measurements.
79
3.3.1.1 Pore Volume Measurements
As mentioned earlier, pore volume can be approximated by the change of volume of brine
in the graduated cylinder before and after brine was drawn into the evacuated cores. Pore
volume of each core, however, can be determined more accurately by two ways,
gravimetric method and tracer tests. In the gravimetric method, the pore volume of each
core was computed based on the weight difference of the core before and after brine was
imbibed into the dry and evacuated core. In the tracer tests, concentration of tracer
solution is analyzed; equal area technique is utilized to calculate pore volume of each core.
Both techniques are discussed in detail in the following sections.
3.3.1.1.1 Gravimetric Method
The core holder was weighed before and after fully saturated with brine. Pore volume of
each core was calculated as follows:
Ws − Wd
PV =
ρ
Where,
PV = pore volume (cc)
Ws = weight of the core holder when it is 100% saturated with brine (g)
Wd = weight of the core holder when it is dry (g)
80
ρ = density of brine at the testing temperature (g/cc)
Core porosities were calculated as follows:
PV
φ=
BV
Where,
Φ= porosity
PV = pore volume (cc)
BV = bulk volume (cc)
3.3.1.1.2 Tracer Tests
Tracer tests were performed to confirm the pore volume of a core. A tracer is an injected
substance that is both measurable (e.g. through UV-visible detection) and conserved (i.e.
not retained, destroyed or created by the core). 1 wt% MgNO3 was used as a tracer in this
study. If tracer concentration is recorded at the core outlet and plotted as a function of
time, an S-shaped curve will form. The S-shaped curve is caused by dispersion of one fluid
into another; otherwise, there would be an instantaneous change in tracer concentration at
breakthrough. Tracer data is typically normalized to scale the tracer concentration values
from zero to one. A typical normalized tracer curve is shown in Figure 3.2.
81
1.00
A2
0.80
Normalized Concentration
0.60
0.40
0.20
teq
A1
0.00
0 200 400 600 800 1000 1200 1400
Time, sec
Figure 3-2 Equal area technique in calculating PV of a core from tracer tests
Normalized concentration is calculated as follows:
CB* − CBo
CB =
CBi − CBo
Where,
CB = normalized tracer concentration
CB*= measured tracer concentration
CBo= initial concentration of brine
CBi = maximum tracer concentration
82
Equal-area technique is utilized to compute PV of a core as follows:
PV = ((teq − to )× q ) − Vd
Where,
Vd = any dead volume (cc)
q = tracer flowrate (cc/min)
to = time at which tracer injection starts (min)
teq= time at which equal area under each side of the curve reaches, A1 = A2
Areas on each side of the tracer concentration, A1 & A2, were calculated using trapezoidal
rule. teq is at A1 = A2.
Several primary assumptions are made in tracer testing [3]
• Fluid B is displacing Fluid A, and the two fluids are miscible
• There is no viscous fingering of one fluid into another, only dispersion
• Flow is single phase
• Fluids are incompressible
• Fluids are of equal density
• Flow is in only one direction
• Fluid velocity is constant
• Flow is through a porous medium of constant porosity and cross-sectional area
83
Figure 3-3 Schematic of tracer setup
Figure 3.3 shows the schematic of the tracer setup. The injection system consists of an
ISCO, Inc. 260DM syringe pump (for mineral oil transfer and injection) and two transfer
cylinders (for tracer solution/brine/mineral oil storage). 1 wt% MgNO3 was used as a
tracer. The inlet of the core holder is connected to the injection system. The three-way
valve allows switching the flow between tracer solution and brine. The outlet of the core is
connected to an UV-visible detector. The UV-visible detector should be turned on and set at
a wavelength of 302 nm at least an hour prior tracer tests. The flow cell of the detector
should be cleaned by methylene chloride, followed by acetone and finally brine injection.
84
Brine passed through core at a constant flow rate until UV-visible detector read a steady
concentration value. At this point, the detector was zeroed and the three-way valve was
switched to tracer injection. The switchover time needed to be recorded manually.
When tracer concentration did not change over time, the valve was switched back to non-
tracer injection to displace the tracer solution from the core. Again, the switchover time
needed to be recorded manually.
3.3.1.2 Permeability Measurements
Pressure drop was recorded during the brine flow tests and used to calculate permeability
of each core based on Darcy’s Law:
 µ  L 
k = q  
 A  ∆P 
Where,
q = flow rate of fluid through the core (cc/sec)
k = permeability (Darcy)
A= area (cc)
µ = viscosity of the fluid (cp)
∆P = pressure drop (atm)
L = length (cm)
85
Brine flow rates were varied to confirm the permeability of each core.
3.3.2 Core Floods
3.3.2.1 Experimental Setup
DPT
Brine
Quizix Pump
CO2
ISCO Pump
BPR
Gas
Test Oil
Oil
ISCO Pump
Figure 3-4 Schematic of core flood setup
The core flood setup is shown in Figure 3.4. The core flood displacement setup consists of a
core holder, injection system, a high-pressure densitometer, a production system and a
computerized data acquisition system. The temperature of the system is maintained in a
Linberg/Blue M oven with Eurotherm temperature control.
86
The injection system consists of a Parker transfer cylinder (oil storage), two Isco, Inc.
260DM syringe pumps (oil/CO2 transfer and injection) and a Quizix Pump (brine transfer
and injection). The capacity of the transfer cylinder is 485 cc. The cylinder can withstand a
maximum pressure of 3000 psi.
The production system consists of a back-pressure regulator, connected to the outlet of the
core holder, to set/control the system pressure. Back pressure regulator models BPR-50 is
a dome-load type, which controls the upstream back pressure to whatever pressure is
applied to the dome. It is designed to operate using compressed gas in the dome and water,
oil, gas in the body. A high pressure bottle of inert gas, such as nitrogen, is required to
pressurize the unit. The back pressure regulator has a working pressure of 5000 psi at
200oF (93oC).
During the experiment, the core effluent was flashed to atmospheric pressure at the outlet
of the back-pressure regulator. The separator fluid was collected in glassware designed for
different stages of displacement. The amount of fluids produced was determined
volumetrically and/or gravimetrically.
Three Valydine pressure transducers are installed to measure pressures at different
locations, such as pressure drop across the core, upstream pressures (CO2/oil/brine
pressure), and downstream pressure (back-pressure regulator pressure). The transducers
measure pressures up to 2500 psi with the accuracy of 0.25% of their full scale (0-2500
psi).
87
3.3.2.2 Experimental Procedures
Two sets of core flood experiments were performed, secondary CO2 flooding and tertiary
CO2 flooding to investigate the effect of operating pressure, in the near-miscible range on
the recovery efficiency and to investigate the effect of mobile water on the recovery
efficiency. In each set of experiment, core flood tests were carried out for a range of
pressures and at the reservoir temperature of 110oF. Crude oil/brine/CO2 injection
volumes were kept the same to compare the amount of oil recovered in each run.
3.3.2.3 Displacement Rate Selection
Residual oil saturation in laboratory core floods has been observed to depend on CO2 flow
rate. Poorer oil recoveries were found at increasing frontal advance rates by Shelton and
Schneider [20]. Low flow rate allows CO2 to diffuse through the network of pores and hence
increase oil recovery by swelling and/or vaporizing. Crude oil/brine/CO2 displacement
rates were selected 0.1cc/min to keep the capillary number as low as 10-6 and to achieve
realistic unit displacement efficiency in laboratory CO2 displacement.
88
3.4 Results and Discussions
3.4.1 Core Characterization Results
Table 3.2 presents the properties of Berea sandstone, Baker dolomite and Arbuckle
dolomite. Bulk volumes of the cores were calculated from at least three measurements of
diameters and lengths. Pore volume of each core was averaged from the results of tracer
tests and gravimetric methods, and from which porosity of each core was calculated.
Permeability to brine was calculated using different flow rates and averaged.
Type Berea sandstone Berea sandstone Arbuckle Baker

A1 A2 dolomite dolomite
Length (cm) 5.86 14.67 5.97 8.07
Cross section 2.53 3.86 2.46 2.34
(cm)
Area (cm2) 5.01 11.67 4.75 4.30
Pore volume (cc) 5.80 34.72 6.05 7.20
Porosity 19.7 % 20.28% 21.3% 20.7%
Permeability 238.5 369.04 2.5 89.7
(mD)
Table 3.2 Core properties
3.4.2 Core Floods Results
3.4.2.1 Secondary CO2 Flooding
89
Berea sandstone core A1 was completely saturated with crude oil prior to CO2 flooding. The
amount of oil recovered by CO2 flooding was determined by weight. Percentage of oil
recovery in core flow tests was determined the same way as in slim tube tests, except pore
volume of the core was used instead of pore volume of the slim tube. % Oil recovery @ 6
PV of CO2 injected was plotted as a function of pressure in Figure 3.5 and was compared
with % oil recovery in slim tube tests @ 1.2 PV CO2 injected at 110oF.
120
100
Recovery of ROIP (%)
80
60
40
20 Slim tube
Berea Sandstone
0
900 1000 1100 1200 1300 1400 1500 1600
Test Pressure, psia
Figure 3-5 Percent of oil recovery @ 6 PV of CO2 injected in core flow tests and percent of oil recovery @ 1.2
PV of CO2 injected in slim tube tests at 110oF
As expected, the oil recovery efficiency in core flow tests was much less than in slim tube
tests. Oil recovery efficiency in core flow tests was approximately 35% of original oil in
place compared with 83 % of original oil in place in slim tube results at 110oF and 1150
psig. The reasons for the difference lie in viscous fingering, channeling and/or bypassing of
90
oil due to core heterogeneities in a core displacement test. Another theory might be the
core was too short for the multiple contact miscibility to develop. Rathmell et al. [21]
reported that the average residual oil for the entire core was observed to decrease with
increasing core length and suggested that up to several feet were required to generate the
miscible transition zone before oil recovery representative of CO2 miscible displacement
was achieved. Therefore, further experiments with longer cores need to be performed to
verify this.
3.4.2.2 Tertiary CO2 flooding
Three groups of rock samples were used in this series of test. Berea sandstone, Arbuckle
dolomite and Baker dolomite were first saturated with brine, followed by crude oil
injection to residual water saturation and brine injection to residual oil saturation prior to
CO2 flooding. At least 10 PV of crude oil/brine injection was used to establish a steady
residual fluid saturation. The amount of oil recovered by CO2 flooding was determined
volumetrically. Table 3.3 to Table 3.6 summarizes CO2 tertiary core flood results for Berea
sandstone/ Arbuckle dolomite/ Baker dolomite at 110oF. Calculations of residual water
saturation (Swr), residual oil saturation due to water-flooding (Sorw), residual oil saturation
due to CO2 flooding (Sorm), and residual water saturation due to CO2 flooding (Swf) are based
on material balance. A sample is attached in the Appendix.
Pressure Swr Sorw Sorm Swf Recovery 1-(Sorm/Sorw)

(psig)
91
905 0.32 0.48 0.31 0.37 35.71
1104 0.32 0.50 0.29 0.39 41.38
1198 0.32 0.48 0.26 0.41 46.43
1317 0.32 0.50 0.21 0.34 58.62
1413 0.32 0.48 0.21 0.34 57.14
Table 3.3 Tertiary CO2 flood results in Ogallah/Berea sandstone A1 at 110oF

(psig)
900 0.32 0.38 0.31 0.43 20.45
1099 (Run 1) 0.35 0.35 0.26 0.46 27.87
1100 (Run2) 0.36 0.36 0.26 0.41 28.13
1206 0.33 0.36 0.20 0.44 45.97
1314 0.36 0.33 0.13 0.51 60.18
Table 3.4 Tertiary CO2 flood results in Ogallah/Berea sandstone A2 at 110oF

(psig)
901 0.38 0.41 0.17 0.51 60.00
1100 0.38 0.41 0.17 0.55 60.00
1200 0.45 0.33 0.08 0.64 75.00
1305 0.45 0.33 0.07 0.64 80.00
1407 0.38 0.38 0.10 0.53 73.91
Table 3.5 Tertiary CO2 flood results in Ogallah/Arbuckle dolomite at 110oF

(psig)
905 0.28 0.39 0.15 0.44 60.71
1109 0.31 0.38 0.13 0.41 66.67
1201 0.34 0.35 0.10 0.45 72.00
1303 0.37 0.35 0.07 0.53 80.00
1402 0.37 0.32 0.07 0.47 78.26
Table 3.6 Tertiary CO2 flood results in Ogallah/ Baker dolomite at 110oF
Figure 3.6 compares percent of oil recovery @ 6 PV of CO2 injected as a function of pressure
for Berea sandstone A1, Baker dolomite and Arbuckle dolomite at 110oF. The recovery
efficiency was similar between two dolomite cores but substantially higher than that in
92
Berea core. These recoveries varied from 65% to 80% for dolomite cores and lesser from
45% to 60 % for sandstone core in the near-miscible region.
100
80
60
40
Berea Sandstone
20 Aruckle dolomite
Baker dolomite
0
900 1000 1100 1200 1300 1400 1500 1600
Pressure (psia)
Figure 3-6 Percent of oil recovery @ 6 PV of CO2 injected as a function of pressure for Berea sandstone
A1, Baker dolomite and Arbuckle dolomite at 110oF
Wylie and Mohanty [22] carried out coreflood studies to examine oil recoveries for first
contact miscible, multiple-contact miscible and immiscible mass transfer processes by
observing recoveries at various wetting states and core orientations and concluded that
mass transfer is enhanced for oil-wet condition compared to water wet condition. Although
wettability of the cores was not examined in this study, it is generally believed that Berea
sandstone is strongly water-wet and to a lesser extent in dolomite cores. After CO2
breakthrough from the core, the extraction or the mass transfer between the bypassed
93
region and flowing CO2 becomes more important to extract the remaining oil inside the
core. The findings from Wylie’s study may explain why the recovery efficiency is higher in
dolomite cores than in sandstone cores for this particular experiment.
As mentioned earlier, crude oil/brine/CO2 displacement rates were selected 0.1cc/min to
keep the capillary number as low as 10-6 and to achieve realistic unit displacement
efficiency in laboratory CO2 displacement. However, since the residual oil saturation in
Berea sandstone A1 was as high as 50%, different displacement rate was selected to drive
the residual oil saturation further down. Crude oil flow rate was selected 1cc/min for the
first 2 PV and 2cc/min for another 4PV during brine displacement test by crude oil. Brine
flow rate was selected 1cc/min for 6 PV during crude oil displacement test by brine. High
pressure drop drove the waterflood residual oil saturation from 0.48-0.50 for Berea
sandstone A1 to 0.33-0.38 for Berea sandstone A2. CO2 flow rate was selected 0.1cc/min
for 6 PV. At current reservoir pressure 1150 psig, the recovery efficiency of Berea
sandstone A1 was approximately 45% whereas the recovery efficiency of Berea sandstone
A2 was 35%. Core flow test was repeated at 1100psig & 110oF for Berea sandstone A2.
Core flow tests at 1100psig and 110oF were reproducible. It is not possible to compare the
recovery efficiency in Berea long core and Berea short core since the pore volume and the
residual oil saturation after waterflooding of two cores were different.
94
70
60
% Oil Recovery, 1- (Sorm/Sorw)
50
40
30
Berea Short Core

20
Berea Long Core-Run 1
10
Berea Long Core - Run 2
0
800 900 1000 1100 1200 1300 1400
System Pressure, psi
Figure 3-7 Recovery efficiency of Berea sandstones A1 & A2 at 1100 psig & 110oF.
3.4.2.3 Effect of Water Saturation on Oil Recovery Efficiency
The residual oil phase after waterflooding and before initiating tertiary CO2 injection exists
in the porous media as a trapped, disconnected phase surrounded by water. High mobile
water saturations affect the CO2 displacement efficiency in an unfavorable way because
water prevents CO2 from contacting and mobilizing the oil, however to what extent is still
unclear.
95
Shelton and Schneider [20] studied the effect of mobile water saturations on oil recovery in
CO2 miscible displacement tests on sandstone cores and concluded that miscible
displacement performances for secondary and tertiary conditions are equivalent.
Tiffin and Yellig [23] evaluated continuous CO2 injection in secondary and tertiary
displacement modes using water-wet and oil wet cores and concluded that mobile water
from a previous waterflooding did not significantly change either the overall recovery or
the mechanism by which miscibility was developed from what was observed in secondary
CO2 displacement tests.
In another study by Campbell and Orr [24], however, visual observations of pore-level
displacement events indicate that CO2 displaced oil much more efficiently in both first
contact and multiple contact miscible displacements when water was absent. Water from a
prior water flood restricted access of CO2 to the oil. The low viscosity of CO2 aggravated
effects of high water saturations because CO2 did not displace water efficiently. Given
enough time, CO2 did however contact trapped oil by diffusing through water to reach, to
swell and to reconnect isolated droplets.
Also, studies by Stalkup [25] have indicated that higher water saturations increase both
the amount of bypassed oil and the longitudinal dispersion coefficient. He also presented
that the dominant mechanisms for recovery of the water blocked oil are miscible flushing
by solvent and recovery by molecular diffusion. Furthermore, Stalkup’s results indicate
96
that the effect of oil trapping by water will be more significant in laboratory core
experiments than in field scale operations.
Shyeh-Yung et al. [7] studied the effect of initial water saturation by comparing secondary
CO2 floods with tertiary CO2 floods. Initial water saturation was about 40% for secondary
CO2 floods and a water-flood preceding a tertiary CO2 floods usually raised the water
saturation to about 75%. Secondary CO2 floods showed more oil recovery than CO2 tertiary
floods at similar pressure and CO2 throughputs, as see in Figure 3.8.
Figure 3-8 Effect of initial water saturation on CO2 flood oil recovery performance by comparing secondary
and tertiary CO2 floods [7]
She suggested several explanations, for instance (1) higher initial oil saturation in
secondary CO2 flood results in oil being more connected (2) lower water saturation in a
secondary flood facilitates mass transport of CO2 to crude oil, allowing more extraction and
displacement of oil (3) intermediate and heavy components extracted from crude oil will
97
make their way to the flowing CO2 phase and be produced more easily in a lower water
saturation secondary flood than in a tertiary flood.
100
Berea sandstone @ Soi
80
Berea sandstone @Sorw
60
40
20
0
900 1,000 1,100 1,200 1,300 1,400 1,500
Pressure (psig)
Figure 3-9 Effect of water saturation on oil recovery efficiency at 110oF
Results from secondary CO2 floods and tertiary CO2 floods using Berea sandstone core
samples are presented in Figure 3.9. Oil recovery efficiency in tertiary CO2 floods is higher
than in secondary CO2 floods. The data suggests that the existence of water phase in the
core is not necessarily detrimental to CO2 displacement efficiency as reported earlier. The
relative permeability of CO2 in the presence of water phase might be reduced, as reported
by Shyeh-Yung et al. [7], which coupled with the reduction of the oil viscosity reduces the
mobility ratio between the oil and CO2 and therefore the recovery efficiency is improved.
Shyeh-Yung measured CO2-rich phase relative permeability (krCO2) in San Andres mixed
wet outcrop cores and observed that secondary CO2 floods had greater end-point solvent
98
mobilities than tertiary CO2 floods (presence of mobile water phase) carried out at the
same temperature. The highest value for krCO2 corresponds to the situation where no
residual oil is present. As residual oil increases, krCO2 decreases. CO2 flood end-point
permeability was plotted as a function of water saturation and residual oil saturation in
Figure 3.10.
Figure 3-10 CO2 flood end-point permeability as a function of water saturation and residual oil saturation in
a mixed wet carbonate [7]
3.5 Conclusions
• Tertiary oil recovery efficiency varied among rock types, from 65% to 80% for
dolomite cores and lesser from 45% to 60 % for sandstone core in the near-miscible
region, from 1100 psig to 1350 psig & at 110oF.
99
• Recovery of more than 60% of the waterflood residual oil was obtained using
Arbuckle reservoir core when CO2 was injected at the current average reservoir
pressure 1150 psig & 110oF.
• Experimental works showed that the presence of water phase improved the relative
permeability of CO2, which coupled with the reduction of the oil viscosity reduced
the mobility ratio between the oil and CO2 and therefore the recovery efficiency was
improved
100
4 Conclusions & Recommendations
4.1 Conclusions
1. Slim tube results indicate that Arbuckle reservoirs is not a suitable candidate for
miscible CO2 flooding, since the MMP for this oil/CO2 system is 1350 psig at 110oF
whereas the operating pressure of the reservoir is 1150 psig.
2. The near miscible range is defined from 1100 psig to 1350 psig at 110oF in our
study.
3. Initial laboratory works showed that at least 80% oil recovery efficiency was
observed in slim tube tests and more than 60% of the waterflood residual oil was
obtained using Arbuckle reservoir core when CO2 was injected at the current
average reservoir pressure 1150 psig & 110oF.
4. Tertiary oil recovery efficiency varied from 65% to 80% for dolomite cores and
lesser from 45% to 60 % for sandstone core in the near-miscible region.
5. The principal oil recovery mechanism in the near-miscible range appeared to be
extraction/vaporization of hydrocarbon components from crude oil into the CO2
rich vapor phase, coupled with enhanced mobility control due to the reduction of oil
viscosity. This suggested that application of carbon dioxide in the field would
require injection and recycling of large volumes of carbon dioxide. Further study is
needed to determine if such a process is economically feasible. However the
101
prospect of recovering up to 1 billion barrels of oil from Arbuckle reservoirs offers
significant economic potential.
6. A representative phase behavior model was developed to describe the fluid
properties and fluid-fluid interaction and would be used in a compositional
simulator to evaluate the potential field application under near-miscible conditions.
4.2 Recommendations
1. Volumetric glassware read as accurately as 0.05-0.1cc, which introduces a large
uncertainty especially when the pore volumes of short cores are approximately 6-7
cc. Either long core samples should be used or volumetric glassware with smaller
divisions should be used.
2. Core flow tests were performed only once for each core and at each test condition.
Core flow tests need to be repeated to see whether the results are reproducible.
3. Wettability test should be performed for Arbuckle dolomite, Baker dolomite and
Berea sandstone. Wettability test could help explaining the difference in tertiary oil
recovery efficiencies among rock types.
4. The possible improvement of CO2 relative permeability in the presence of water
phase was used to explain for the better oil recovery efficiency seen in tertiary CO2
floods than the oil recovery efficiency seen in secondary CO2 floods. Experimental
102
measurements of CO2 relative permeability in the presence of Sorm should be
performed to verify this.
103
REFERENCES
1. Stosur, G.J., et al., Enhanced Oil Recovery in North America: Status and Prospects.
Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 1990. 12(4):
p. 429 - 437.
2. Holm, L.W., Miscibility and miscible displacement. Journal Name: J. Pet. Technol.;
(United States); Journal Volume: 38:9, 1986: p. Medium: X; Size: Pages: 817-822.
3. Green, D.W. and G.P. Willhite, Enhanced Oil Recovery. 1998: SPE.
4. Holm, L.W. and V.A. Josendal, Effect of Oil Composition on Miscible-Type
Displacement by Carbon Dioxide. 1982. 22(1): p. 87-98.
5. Metcalfe, R.S. and L. Yarborough, The Effect of Phase Equilibria on the CO2
Displacement Mechanism. 1979. 19(4): p. 242-252.
6. Klins, M.A., Carbon Dioxide Flooding Boston: International Human Resources
Development Corporation.
7. Shyeh-Yung, J.-G.J., Mechanisms of Miscible Oil Recovery: Effects of Pressure on
Miscible and Near-Miscible Displacements of Oil by Carbon Dioxide, in SPE Annual
Technical Conference and Exhibition. 1991, 1991 Copyright 1991, Society of Petroleum
Engineers Inc.: Dallas, Texas.
8. Schechter, D.S., et al., Wellman Unit CO2 Flood: Reservoir Pressure Reduction and
Flooding the Water/Oil, in SPE Annual Technical Conference and Exhibition. 1998,
Society of Petroleum Engineers: New Orleans, Louisiana.
9. Dong, M., S.S. Huang, and R. Srivastava, A Laboratory Study on Near-Miscible CO
Injection in Steelman Reservoir. 2001. 40(2).
104
10. Hadlow, R.E., Update of Industry Experience With CO2 Injection, in SPE Annual
Technical Conference and Exhibition. 1992, 1992 Copyright 1992, Society of Petroleum
Engineers Inc.: Washington, D.C.
11. Grigg, R.B., M.D. Gregory, and J.D. Purkaple, The Effect of Pressure on Improved
Oilflood Recovery from Tertiary Gas Injection. SPE Reservoir Engineering, 1997. 12(3):
p. 179-188.
12. Rahmatabadi, K.A., The role of interfacial tension in near-miscible CO2 injection, in
Chemical and Petroleum Engineering. 2006, University of Kansas: Lawrence. p. 107.
13. Ren, W. and A.M. Scurto, High-pressure phase equilibria with compressed gases (vol 78,
art no 125104, 2007). Review of Scientific Instruments, 2008. 79(2).
14. Nagarajan, N. and R.L. Robinson, Equilibrium phase compositions, phase densities, and
interfacial tensions for carbon dioxide + hydrocarbon systems. 2. Carbon dioxide + n-
decane. Journal of Chemical & Engineering Data, 1986. 31(2): p. 168-171.
15. Jennings, D.W. and R.C. Schucker, Comparison of High-Pressure Vapor−Liquid
Equilibria of Mixtures of CO2 or Propane with Nonane and C9 Alkylbenzenes. Journal of
Chemical & Engineering Data, 1996. 41(4): p. 831-838.
16. Hand, J.L. and W.V. Pinczewski, Interpretation of Swelling/Extraction Tests. SPE
Reservoir Engineering, 1990. 5(4): p. 595-600.
17. Ahosseini, A. and A. Scurto, Viscosity of Imidazolium-Based Ionic Liquids at Elevated
Pressures: Cation and Anion Effects. International Journal of Thermophysics, 2008.
29(4): p. 1222-1243.
105
18. Negahban, S. and V.J. Kremesec Jr., Development and Validation of Equation-of-State
Fluid Descriptions for CO2/Reservoir-Oil Systems. SPE Reservoir Engineering, 1992.
7(3): p. 363-368.
19. Orr Jr., F.M., et al., Laboratory Experiments To Evaluate Field Prospects for CO2
Flooding. SPE Journal of Petroleum Technology, 1982. 34(4): p. 888-898.
20. SHELTON, J.L., The Effects of Water Injection on Miscible Flooding Methods Using
Hydrocarbons and Carbon Dioxide. 1975. 15(3): p. 217-226.
21. Rathmell, J.J., F.I. Stalkup, and R.C. Hassinger, A Laboratory Investigation of Miscible
Displacement by Carbon Dioxide, in Fall Meeting of the Society of Petroleum Engineers
of AIME. 1971, 1971 Copyright 1971: New Orleans, Louisiana.
22. Wylie, P.L. and K.K. Mohanty, Effect of Wettability on Oil Recovery by Near-Miscible
Gas Injection. SPE Reservoir Evaluation & Engineering, 1999. 2(6): p. 558-564.
23. Tiffin, D.L. and W.F. Yellig, Effects of Mobile Water on Multiple-Contact Miscible Gas
Displacements. 1983. 23(3): p. 447-455.
24. Campbell, B.T. and F.M.O. Jr., Flow Visualization for CO2/Crude-Oil Displacements.
1985. 25(5): p. 665-678.
25. Stalkup, F.I., Displacement of oil by Solvent at High Water Saturation. 1970. 10(4): p.
337-348.
106
APPENDICES
Calculations of Swr, Sorw, Sorm and Swf are based on material balance. Following is an example
of how to calculate those values for Berea Sandstone at 1104 psig & 110oF.
Dead volume = 2.25 cc
Average pore volume of Berea Sandstone = 5.80 cc
Calculate residual water saturation (Swr) at the end of brine displacement test by crude oil
Assuming that the core was fully saturated with brine and the line was also completely
filled with brine, volume of brine in the core and in the line = 2.25 + 5.80 = 8.05 cc
Volume of brine collected at the end of the displacement test = 6.20 cc
Volume of brine in the core at the end of the displacement test = 8.05- 6.20 = 1.85 cc
Volume of oil in the core at the end of the displacement test = 5.80 – 1.85 = 3.95 cc
Swr = 1.85/8.50 = 0.32
So = 3.95/8.50 = 0.68
Calculate residual oil saturation (Sorw) at the end of crude oil displacement by brine
Since the line was completely filled with crude oil, volume of crude oil in the core and in the
line = 3.95 + 2.25 = 6.20 cc
107
Volume of crude oil collected at the end of the displacement test = 3.30 cc
Volume of crude oil in the core at the end of the displacement test = 6.20 – 3.30 = 2.90 cc
Volume of brine in the core at the end of the displacement test = 5.80 – 2.90 = 2.90
Sorw = 2.90/5.80 = 0.50
Sw = 2.90/5.80 = 0.50
Calculate residual oil saturation (Sorm) at the end of CO2 displacement test
Volume of crude oil prior to CO2 displacement test = 2.90 cc
Volume of crude oil collected at the end of the displacement test = 1.20 cc
Sorm = 1.70/5.80 = 0.29
% Oil Recovery = (0.50-0.29)*100%/0.50 = 41.4 %
Volume of brine prior to CO2 displacement test = 2.90 + 2.25 = 5.15 cc
Volume of brine collected at the end of the displacement test = 2.90 cc
Swf = 2.25/5.80 = 0.39
108

Bui Ku 0099M 11098 DATA 1

Uploaded by

Copyright:

Available Formats

Bui Ku 0099M 11098 DATA 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bui Ku 0099M 11098 DATA 1

Uploaded by

Copyright:

Available Formats

Near-Miscible CO2 Application to Improve Oil Recovery

the degree of Master of Science

Jyun Syung Tsau

Date defended: __________________________

that this is the approved version of the following thesis

Near-Miscible CO2 Application to Improve Oil Recovery

Jyun Syung Tsau

Date approved: __________________________

injection is normally operated at a pressure above the minimum miscibility pressure

application and improve our understanding of the mechanisms of near-miscible CO2

flooding by conducting appropriate experimental work and reservoir simulation. The

near-miscible conditions. The suite of laboratory experiments used to evaluate the

mechanism in the near-miscible range appeared to be extraction/vaporization of

volumes of carbon dioxide. Further study is needed to determine if such a process is

Arbuckle reservoirs offers significant economic potential.

TABLE OF CONTENTS .......................................................................................................................... IV

NOMENCLATURE ................................................................................................................................ XII

ACKNOWLEDGMENTS ...................................................................................................................... XIV

1 INTRODUCTION AND LITERATURE REVIEW ....................................................................... 1

1.1 THE BASICS OF CO2 EOR.......................................................................................................................... 4

1.1.1 CO2 Properties......................................................................................................................................... 4

1.1.2 Immiscibility & Miscibility................................................................................................................. 6

1.1.3 Minimum Miscibility Pressure ......................................................................................................... 8

1.1.4 Mechanisms of Oil Displacement by CO2 ..................................................................................... 9

1.1.5 Mechanisms for CO2 Miscibility with Oil .................................................................................. 11

1.2 MOTIVATIONS BEHIND THIS STUDY....................................................................................................... 15

1.3 OBJECTIVES OF THE STUDY ..................................................................................................................... 21

2 PHASE BEHAVIOR STUDIES ..................................................................................................... 23

2.1 FLUID PROPERTIES .................................................................................................................................. 23

2.2 SLIM TUBE DISPLACEMENTS ................................................................................................................. 26

2.2.1 Experimental Setup and Specifications .................................................................................... 27

2.2.2 Experimental Procedures ............................................................................................................... 29

2.2.3 Results and Discussions ................................................................................................................... 31

2.3 SWELLING/EXTRACTION TESTS ............................................................................................................ 37

2.3.1 Experimental Setup and specifications .................................................................................... 37

2.3.2 Experimental Procedures ............................................................................................................... 39

2.3.3 Experimental Principles .................................................................................................................. 41

2.3.4 Apparatus Validation ....................................................................................................................... 43

2.3.5 Results and Discussions ................................................................................................................... 44

2.3.5.1 Effect of System Pressure ........................................................................................................................ 44

2.3.5.2 Effect of System Temperature ................................................................................................................. 47

2.3.5.3 Effect of Initial Oil Volume ....................................................................................................................... 50

2.3.6 Conclusions ........................................................................................................................................... 55

2.4 VISCOSITY MEASUREMENTS................................................................................................................... 55

2.4.1 Principle of Operation ...................................................................................................................... 56

2.4.2 Experimental Setup and Specifications .................................................................................... 57

2.4.3 Experimental Procedures ............................................................................................................... 59

2.4.3.1 Oil Viscosity Measurement Procedure .................................................................................................... 59

2.4.3.2 Oil/CO2 Mixture Viscosity Measurement Procedure ............................................................................. 61

2.4.4 Results and Discussions ................................................................................................................... 62

2.4.5 Conclusions ........................................................................................................................................... 63

2.5 PHASE BEHAVIOR MODEL ...................................................................................................................... 63

2.5.1 Phase Behavior Modeling using WINPROP ............................................................................ 65

2.5.2 Equation of State Characterization ........................................................................................... 68

2.5.3 Slim Tube Modeling using GEM ................................................................................................... 70

3 CORE FLOW TESTS ..................................................................................................................... 76

3.2 FLUIDS ....................................................................................................................................................... 79

3.3 EQUIPMENTS AND PROCEDURES ........................................................................................................... 79

3.3.1 Core Characterization ..................................................................................................................... 79