REVIEW OF LITERATUR.....

review_12.pdf

Home >English homework help >REVIEW OF LITERATUR.....

SPE 146649 Impact of Asphaltene Nanoscience on Understanding Oilfield Reservoirs Oliver C. Mullins,1 A. Ballard Andrews,1 Andrew E. Pomerantz,1 Chengli Dong,1 Julian Y. Zuo,1 Thomas Pfeiffer,1 Ahmad S. Latifzai,1 Hani Elshahawi,2 Loïc Barré,3 Steve Larter4 1. Schlumberger Oilfield Services, 2. Shell Exploration and Production Company, Inc, 3. IFP Energies Nouvelles , 4. PRG, University of Calgary & Gushor Inc.

Copyright 2011, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 30 October–2 November 2011. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract Understanding asphaltene gradients and dynamics of fluids in reservoirs had been greatly hindered by the lack of knowledge of asphaltene nanoscience. Gravitational segregation effects on oil composition, so important in reservoir fluids, are unresolvable without knowledge of (asphaltene) particle size in crude oils. Recently, the “modified Yen model” also known as the Yen-Mullins model, has been proposed describing the dominant forms of asphaltenes in crude oils: molecules, nanoaggregates and clusters. This asphaltene nanoscience approach enables development of the first predictive equation of state for asphaltene compositional gradients in reservoirs, the Flory-Huggins-Zuo (FHZ) EoS. This new asphaltene EoS is readily exploited with “downhole fluid analysis” (DFA) on wireline formation testers thereby elucidating important fluid and reservoir complexities.

Field studies confirm the applicability of this scientific formalism and DFA technology for evaluating reservoir compartmentalization and especially connectivity issues providing orders of magnitude improvement over tradional static pressure surveys. Moreover, the mechanism of tar mat formation, a long standing puzzle, is largely resolved by our new asphaltene nanoscience model as shown in field studies. In addition, oil columns possessing large disequilibrium gradients of asphaltenes are shown to be amenable to the new FHZ EoS in a straightforward manner. We also examine recent developments in asphaltene science. For example, important interfacial properties of asphaltenes have been resolved recently providing a simple framework to address surface science. At long last, the solid asphaltenes (as with hydrocarbon gases and liquids) are treated with a proper chemical construct and theoretical formalism. New asphaltene science coupled with new DFA technology will yield increasingly powerful benefits in the future.

Introduction It is widely acknowledged that reservoir engineering is inextricably linked to the use of cubic equations of state to model compositional gradients and phase behavior. Cubic equations of state are modifications to the van der Waals equation which itself is derived from the ideal gas law. Cubic equations of state are derived to treat gas-liquid equilibria and are not a formalism to treat solids. Hence, they are grossly inadequate to handle molecularly or colloidally suspended solids. Crude oils contain not only gases and liquids but also solids, the asphaltenes. It is not proper to treat the solid asphaltenes with equations derived from the idea gas law. For example, cubic equations of state require knowledge of the critical point, the point at which the liquid and gas properties are identical while asphaltenes have no liquid phase, no gas phase and no critical point. Specifically, there had been no first principles method to model asphaltene gradients in reservoirs. Indeed, this led to a general misunderstanding of black oils. Condensates have relatively high GOR compared to black oils [1] and high GOR fluids generally exhibit large compositional gradients.[1-3] Cubic equations of state yield homogeneous compositions for black oils due to their characteristic low GOR.[1-3] Consequently, there has been the erroneous assumption that black oils are homogeneous because cubic equations of state give this result. Nevertheless, numerous geochemical studies indicate chemical compositional variations do exist laterally and vertically in many black oil reservoirs.[4] As mentioned above, cubic equations of state cannot model asphaltene gradients in any first principle approach. Moreover, black oils are best described by their asphaltene content, not their (low) GOR. Since viscosity depends exponentially on asphaltene content,[5] it

2 SPE 146649

behooves the operator to model asphaltene gradients. We are now able to employ a first-principles model of asphaltene gradients for the first time. The difficulty in petroleum science had been that nobody knew the size of the asphaltene (molecular and colloidal) particles in crude oil. For gravity, size counts. Without known size, the gravity term is unknowable, precluding modeling asphaltenes. Moreover, the use of cubic equations of state for mixtures works well only for nearly ideal systems such as the hydrocarbons. Hydrocarbons are very weakly interacting, for example keeping them in the liquid phase, not solid phase, at room temperature even for large carbon number. However, for acid gas components such as CO2 and H2S and even for H2O, the intermolecular interactions are stronger than for hydrocarbons. For example, H2O is a liquid while much heavier hydrocarbons are gases at room temperature. For treating mixtures with acid gas species and H2O, the cubic EoS methods are much less effective. Extending these same cubic EoS methods to the solid asphaltenes, with their large intermolecular interaction is ill-advised. A novel first-principles approach is needed. Asphaltene Nanoscience After considerable effort by many workers [6], the molecular and colloidal structure of asphaltenes in crude oil and in laboratory solvents has been worked out. Figure 1 shows a representation of asphaltene molecular architecture and two explicit colloidal species. This model was introduced as the modified Yen model [7,8] and has also been called the Yen-Mullins model.[9,10] Of course, asphaltenes contain many different molecules; nonetheless, this paradigm serves our purposes well as we shall see.

Figure 1. Asphaltene nanoscience; the modified Yen model also known as the Yen-Mullins model.[7,8] For low asphaltene concentrations, such as in condensates, asphaltenes are dispersed as a true molecular solution. At somewhat higher concentrations such as in black oils, asphaltenes are dispersed as nanoaggregates. At yet higher concentrations such as in heavy oils, asphaltenes are dispersed as clusters. Diameters in nanometers are listed in the figure. Asphaltene Molecular Structure. There had been an incorrect consensus that asphaltenes consist of large polymeric molecules with many isolated polycyclic

aromatic hydrocarbon ring (PAH) components. (Fig. 1, left, shows a molecule with one central PAH.) The first molecular diffusion measurements of asphaltenes utilizing time resolved fluorescence depolarization (TRFD), yielded two important findings: 1) asphaltene molecules undergo very fast diffusion, they are monomeric in size, not polymeric, 2) small PAH ring systems diffuse ten times faster than large asphaltene PAHs, they are not cross-linked; thus one PAH per molecule.[11] Recent results provide powerful confirmation of both these key results. Two-step laser desorption, laser ionization mass spectrometry (L2MS) has been applied to asphaltenes confirming their low molecular weight.[12] An infrared laser is used to desorb asphaltenes into a vacuum, a UV laser is then used to ionize the asphaltenes, and time of flight is used to measure molecular weight. By increasing the UV laser pulse energy, the resulting ions can be fragmented.[9] Figure 2 shows the result of fragmenting asphaltenes and various model compounds in a unimolecular process.[9]

Asphaltenes

Monomer Model Compounds

Polymer Model Compounds

Asphaltenes

Figure 2. Two-step laser desorption, laser ionization mass spectrometry of asphaltenes and model compounds. At higher UV laser pulse energy, polymers fragment; asphaltenes and monomers do not fragment confirming the monomer (one PAH per molecule) molecular architecture of asphaltenes (cf. Fig. 1).[9] Figure 2 shows that polymer model compounds fragment at higher laser power while monomer model compounds and asphaltenes are more stable and resist fragmentation.

SPE 146649 3

Asphaltenes survive for geologic time, stability against fragmentation is expected. Asphaltene molecules are confirmed to be predominantly monomers, that is, one PAH per molecule (often with heteroatoms) as depicted in Fig. 1. Asphaltene Nanoaggregates. Fluorescence measurements determine that asphaltene molecules start to associate at low concentrations (~50 mg/liter in toluene).[13] Many different methods have been used to measure the asphaltene critical nanoaggregate concentration (CNAC) including high-Q ultrasonics [14], NMR H-Index [15], NMR diffusion [15], AC- Conductivity [16], DC-Conductivity [17,18], and centrifugation [19]. All of these measurements are in agreement that nanoaggregate growth terminates at ~100 mg/liter in toluene. The size of nanoaggregates and indeed their internal structure have been delineated in studies where both small angle neutron scattering (SANS) and small angle x-ray scattering (SAXS) are directly compared. As is well known to petrophysicists, neutrons scatter preferentially off hydrogen nuclei which are concentrated in the alkane component of asphaltenes while x-rays scatter off electrons which are concentrated in the carbon rich aromatic (or PAH) component of asphaltenes.[20]

Figure 3. Direct comparison of x-ray scattering (SAXS) and neutron scattering (SANS) absolute cross sections. The divergence in the x-ray versus neutron scattering that occurs at a wave vector q of ~0.07Å-1 (~2.8 nm diameter) is consistent with the nanoaggregate structure and size shown in Fig. 1.

The aromatic rings are stacked in the interior and the alkanes are oriented outwards. Figure 3 is consistent with the picture that electron rich aromatics are concentrated in the nanoaggregate interior while the alkyl substituents are concentrated in the nanoaggregate periphery. The monomer molecular structure naturally dictates this nanoaggregate structure.[14] The PAHs in the molecular interior are attractive while the alkane substituents act to sterically repel other asphaltene molecules. After several asphaltene molecules aggregate, the nanoaggregate periphery is largely covered with repulsive alkane substituents precluding further aggregate growth.[14] More recent analyses of x-ray and neutron scattering data confirms that there is only a single stack of PAHs in the asphaltene nanoaggregate.[21] Again this is consistent with the nanoaggregate in Fig. 1. There are three different types of molecular sizes involved, radius of gyration, relevant for SANS and SAXS measurements, hydrodynamic radius, relevant for diffusion and Stokes drag, and the physical size, relevant for gravitation and centrifugation.[22] Of course, exact agreement among these different sizes is not expected. Asphaltene Clusters. Asphaltene nanoaggregates are the smallest colloidal particle of asphaltenes (molecules are not colloidal), but they are not the only colloidal asphaltene particle. Many systems exhibit colloidal structure, such as micelles of soap in water. However, not all colloidal systems exhibit multiple colloidal particle sizes. A clear demonstration of the formation of asphaltene clusters in toluene was obtained by measurement of the flocculation kinetics of asphaltene flocs upon addition of n-heptane to asphaltene-toluene solutions.[23] Below a concentration of ~3 g/liter, the kinetics of floc formation are diffusion limited aggregation (DLA); whereas above this concentration the kinetics are reaction limited aggregation (RLA). This is consistent with nanoaggregates that are ‘stickier’ whereas the fractal clusters [21] are less adherent. A morphological change of the cluster surface is evidently required for floc formation; this is consistent with RLA.[24]

2 SPE 146649

Figure 4. Critical cluster concentration of asphaltenes in toluene. When clusters form from nanoaggregates, the conductivity per unit mass decreases due to an increase in Stokes drag of the charge carriers. Comparison of the slopes provides an estimate for cluster size to consist of ~8 nanoaggregates.[18] The size of clusters was recently measured using DC- conductivity of asphaltene solutions. Less than 10-4 mole fraction of asphaltene is charged in toluene; charge carriers act as tracers to monitor asphaltene dynamics. The critical clustering concentration is ~2.5 g/liter for asphaltene in toluene.[18] The nanoaggregate is tightly bound and its size is limited by the asphaltene molecular architecture with the attractive PAH on the molecular interior and the sterically repulsive alkane chains on the molecular exterior. The clusters form only at much higher concentration because the attractive forces of one nanoaggregate to another are so much weaker. The fixed size of the clusters found in the lab and the field is due in part to the limited range of asphaltene concentrations that have been probed. At much higher asphaltene concentrations, the heavy oil will not flow easily. The latest experiments in asphaltene science demonstrate repeated consistency with the asphaltene nanoscience model proposed in Figure 1. This model provides a foundation that has far reaching implications in many areas associated with the production of crude oil. The success of such a simple model relies on its correct physics! Asphaltene Interfacial Activity. Asphaltene molecules are of moderate size and, as such, can exhibit a high degree of molecular orientation at the interface. For very small molecules, entropy would presumably disfavor alignment at room temperature, while for very large molecules, entanglement would preclude high degrees of alignment. The first direct measurements of asphaltene molecular alignment in Langmuir-Blodgett (L-B) films of asphaltenes have been performed by the optical method sum frequency generation (SFG).[25] An IR and visible

laser are focused on the same point on an interface. Nonlinear mixing at the interface can result in the creation of sum-frequency (UV) photons. The IR laser wavelength is scanned across molecular vibrational resonances. By controlling the polarization of the two lasers and detected UV beam, molecular orientation is probed. A film of asphaltene on water is created by evaporation a toluene solution of asphaltenes on water. The film is then transferred to a solid substrate for measurement – called a Langmuir-Blodgett film.

Stretch CC aromatic

Bend; ‐CH3, ‐CH2‐ aliphatic

D |

D ||

SSP UG8 Asphaltene

SPS UG8 Asphaltene

SFG on L-B Film

S H

Asphaltene Molecule

Wavenumbers (cm‐1)

Sig na l

Figure 5. Asphaltene molecular orientation as determined by sum frequency generation (SFG),[25] the nonlinear mixing of an IR and visible photon to form a UV photon. By polarization methods, it is shown that the aromatic ring systems of asphaltenes lie in the interfacial plane on the water surface, while the alkyl chains lie perpendicular to this plane. The wettability properties of crude oil have been addressed largely through phenomenological methods. With knowledge of asphaltene molecular architecture contained within the nanoscience model of Fig. 1, it is now possible to better understand detailed properties of asphaltene films from first principles. Linking first principles to surface wetting characterization enables development of methods to alter these properties. Oilfield Applications of Asphaltene Nanoscience Downhole Fluid Analysis (DFA). The use of asphaltene nanoscience for reservoir characterization is predicated on the ability to perform measurements of fluid properties such as asphaltene content and GOR at many points within the reservoir. Measurement of fluid properties at only a few points in the reservoir precludes robust understanding of the primary governing physics controlling fluid distributions in reservoirs. Indeed, the antiquated concept of getting “the oil” using a wireline sampling tool inhibited proper reservoir characterization; this despite numerous examples of fluid complexities.[3,4] Downhole fluid analysis (DFA) is an essential technology for characterizing fluid gradients and for understanding

SPE 146649 5

corresponding implications for reservoir properties. DFA is now routinely performed on wireline formation sampling tools.

Figure 6. Wireline formation sampling tool configured with DFA optical analyzers. DFA helps acquisition of representative samples. DFA also reveals reservoir fluid complexities during the wireline job. Consequently, the wireline measurement program can be altered to match the complexity of the oil column, improving efficiency. To exploit new developments in asphaltene science as well as traditional fluid analyses for reservoir evaluation, it is essential to have adequate data particularly regarding fluid gradients and fluid discontinuities. DFA is the only way to achieve this objective; Fig. 6 shows a typical tool configuration for DFA. DFA reveals fluid complexities during the wireline job, thus, the complexity and cost of the wireline job can be matched to the complexity of the oil column. The operator does not pay for unnecessary analyses. Without DFA, there is no means to reveal fluid complexities in real time; presumptions of fluid simplicity often prevail which are often incorrect, leading to subsequent problems. DFA measurements now include GOR and some hydrocarbon composition, relative asphaltene content, density, viscosity, fluorescence, and CO2. More fluid chemical analytes are being added regularly. Compartmentalization. The costly, bad news of compartmentalization is often easier to uncover with adequate DFA data. Fluid density inversions (higher density fluids higher in the oil column) immediately reveal sealing barriers. Figure 7 shows just such a case. Right above the depth of x400 meters, a low GOR is found. Right below x400 meters a crude oil with ~4 times the GOR is found. GOR is a reasonable proxy for density. In this column, a high density crude oil is found right above a low density crude oil. Clearly a sealing barrier is indicated.

This was delineated with 11 DFA stations which were needed due to the complexity of the crude oil column. High GOR is consistent with large GOR gradients within an equilibrium context.[1,2] In a section below, fluid disequilibrium will also be considered and can contribute significantly to fluid gradients.

GOR (scf/bbl) 1000 2000 3000 4000 5000

x200

x300

x400

x500

x600

x700

x800

Depth (m)

Figure 7. A low GOR (high density) fluid at x350 m is found above a high GOR (low density) fluid at x410m; a sealing barrier is indicated at x400 meters. The complexity of the oil column with highly variable GOR mandates a relatively large number of DFA stations; in turn revealing fluid complexities and a sealing barrier. In some cases, the column does not exhibit a fluid density inversion but does reveal an “asphaltene inversion”. Asphaltenes are more dense than the crude oil. They do not float in oil; if anything, they sink to lower points in the column. In Fig. 8, DFA color inversions (higher color fluid higher in the oil column), thus asphaltene concentration inversions are observed.[26] For example, crude oil A has three times the color magnitude of crude oil B, thus three times the asphaltene content.[27] Thus a sealing barrier is between these zones. However, fluid densities are too similar to make the same assessment. The low asphaltene content (~1%) accounts for the low impact of asphaltenes on overall fluid density.

2 SPE 146649

Color @ 815 nm Fluorescence Intensity

DFA Color

DFA

Fluorescence

Natural Gamma‐Ray

Formation Pressure (psi)

Tr ue

V er ti ca l D

ep th (f ee t)

Figure 8. Compartmentalization (sealing barriers) is shown by asphaltene density inversions, higher asphaltene content (and higher color, less fluorescence) higher in the column. Also, the lack of pressure communication indicates compartmentalization which made this prospect unattractive.[26] The crude oils from the column in Fig. 8 were analyzed by the highest resolution mass spectrometer on earth. Tens of thousands of individual components were resolved. However, no differences were noted from asphaltenes in different zones.[26] The maltenes of these crude oils were analyzed by two-dimensional gas chromatography; again no differences amongst these crude oils were discernable.[28] The chemical composition of the different constituents of these crude oils are the same. It is the concentration that differs and thus concentration is most useful for compartment identification. DFA methods are very sensitive to this concentration.[3] Reservoir Connectivity. It is far more valuable and difficult to verify the good news of reservoir connectivity versus the bad news of compartmentalization. The significant expense of well testing precludes its use in many settings. The difficulty of establishing connectivity causes this attribute often to be the largest risk in development.[3] Moreover, reservoirs that are not conclusively established to be compartmentalized have often been presumed to be connected and thus be large. However, geostatistics teaches that small geophysical objects are always much more numerous than large geophysical objects.[3] Connectivity in reservoirs must be proven, not presumed. Pressure surveys have long been used to address compartmentalization. The statement “if two zones are not in pressure communication, they are not in flow communication,” is true. However, the corresponding statement “if two zones are in pressure communication, they are in flow communication” does not follow from

the first statement and is often incorrect. Indeed, presuming the two logical statements are self consistent results from a fatal flaw in logic (see footnote). Pressure communication can occur on geologic time with minimal flow volumes while flow communication must occur on production time; these constraints differ by six orders of magnitude. Pressure communication can occur through tiny permeability while flow communication requires much larger permeability.[3,29] Pressure communication is a necessary but insufficient condition for flow communication. New methods to assess connectivity are needed. Determination of fluid equilibrium in a reservoir is a stringent method to determine reservoir connectivity. If the fluids are equilibrated throughout a reservoir, the reservoir is very likely connected. This does not imply that if there is disequilibrium, then there is no connectivity; we shall examine disequilibrium in a subsequent section. Reservoir fluids enter the reservoir necessarily out of their final thermodynamic equilibrium. For example, under a normal burial sequence, light fluids enter the reservoir later in time. These lighter fluids must then undergo the equilibration process. Typically, with charging, the fluids exist initially in the reservoir as a stratified sequence from lightest at the top to heaviest at the bottom.[30] The process of equilibration then requires massive fluid flow in the reservoir which takes a long time, especially if by diffusion. If low permeability barriers are present, then this process can be prohibitively slow.

Figure 9. Time scales to achieve fluid equilibrium and pressure equilibrium for a tilted sheet reservoir with a barrier of poor permeability in the middle. Reservoir modeling shows that fluid equilibration is ten million times slower than pressure equilibration.[29] Very good connectivity is needed for fluid equilibration, not for pressure equilibration. Footnote: The logical statement “If A, then B” implies “if not B, then not A”. These statements are related as

SPE 146649 7

contrapositives and follow logically. Assigning flow communication FC=A, and pressure communication PC=B, we have “If FC, then PC” implies “If not PC, then not FC.” This is logically true. However, the statement “if A, then B” does not imply “if not A, then not B.” The 2nd statement is called a nonsequitur. The statements “if not PC, then not FC” and “if PC, then FC” are related as nonsequiturs in logic; it is logically invalid to assume one statement follows from the other. Figure 9 shows that the measurement of fluid equilibrium to evaluate reservoir connectivity is ten million times better then the measurement of pressure equilibration for the same purpose. Obviously, fluid equilibration should now be used to assess reservoir connectivity as a complementary measurement to pressure. To assess fluid equilibrium, equations of state (EoS) are needed. The Peng-Robinson equation is a “cubic equation of state” that is effective for treating many properties of crude oil. The Peng-Robinson equation, developed in 1976, is derived from the original cubic EoS, the van der Waals equation, developed in 1873. The cubic EoS was developed to treat gas-liquid equilibria and can be used on gas-liquid fluid distributions to assess reservoir connectivity. However, it is frequently the case that phase separated gas does not equilibrate well with the liquid column, thereby limiting somewhat the use of the cubic EoS to assess reservoir connectivity. Moreover, for low GOR black oils and heavy oils, there is very little GOR gradient, so measurement of GOR gradients alone provides little insight into reservoir architecture. Especially for these cases, the primary gradient of interest is the asphaltene gradient; crude oils generally contain solids, the asphaltenes, and an appropriate EoS is need. As depicted in Fig. 1, asphaltenes are often colloidal. There is no provision in any cubic EoS to treat colloidal solids. A different equation is needed. The Flory-Huggins-Zuo Equation of State. In a general sense, equations of state treat both phase behavior and gradients. The cubic EoS does just this for gas-liquid equilibria. For many years, the Flory-Huggins equation has been used to treat the phase behavior of asphaltenes.[31] For phase behavior, the solubility of asphaltenes in the liquid phase is key and is represented by the “solubility parameter”, a measure of the volumetric density of chemical intermolecular interaction. This theory quantifies the well known axiom “like dissolves like” through the solubility parameter. In order to extend this equation to treat gradients, it is necessary to include the effect of gravity which depends on the size of the particular asphaltene species in play. This is now known for crude oils and laboratory solvents (cf. Fig. 1). In addition, it is essential to understand how GOR influences the solvation of asphaltenes. With the gravity term included, and proper accounting of the solvation term,[32,33] the Flory-Huggins-Zuo (FHZ) equation results.[34]

( ) ( )

( ) ( ) ( )[ ] ⎟ ⎟

⎠

⎞

⎜ ⎜

⎝

⎛ −−− −⎟

⎠

⎞ ⎜ ⎝

⎛−⎟ ⎠

⎞ ⎜ ⎝

⎛+ −Δ

== RT

v v

RT hhgv

h h

hOD hOD hahaa

a 22

2 12

exp δδδδρ

φ φ

Eq. 1. where OD(hi) is the optical density (measured by DFA) at a particular color channel at height hi in the oil column, φa(hi) is the corresponding asphaltene concentration, va is the molar volume of the asphaltene species (molecule, nanoaggregate or cluster), v is the molar volume of the oil phase, g is earth’s gravity, Δρ is the density difference between asphaltenes and the liquid phase, R is the ideal gas constant, T is temperature, and δa is the solubility parameter of asphaltene, and δ the solubility parameter of the oil. The first term in the exponential includes Archimedes buoyancy of an object in a liquid (perhaps more familiar as mgh). Gravity tends to accumulate the asphaltenes towards the base of the column; this is explicitly counteracted by thermal energy, thus temperature is in the denominator. This gravity term explicitly depends on the molecular weight and colloidal size of the asphaltenes as given in the Yen-Mullins model and is shown in Fig. 1. For larger species such as clusters, the gravity term becomes large giving rise to significant concentrations of asphaltenes towards the base of the oil column. For low GOR black oils and heavy oils, the other terms in Eq. 1 tend to be small, so still the simple gravity term dominates.[33] The second and third terms in Eq. 1 are simply the Flory- Huggins entropy term. Entropy tends to randomize the distribution counteracting any gradients. The entropy term is not very large for asphaltenes in crude oils.[32] The last term in the exponential is the solubility term as discussed above. With the precept “like dissolves like”, this term accounts for decreasing solubility of asphaltenes with increasing chemical difference between the liquid phase and the asphaltenes. In particular, large GOR gives a low density, alkane rich liquid phase that is chemically very different than asphaltenes, thus decreasing asphaltene solubility.[35] If there is no GOR gradient, then there is effectively no difference in the solubility parameter from the top to the bottom in the oil column; then this solubility term does nothing to create an asphaltene gradient. Thus, for low GOR oils where small GOR gradients are the norm, this term is unimportant. In contrast, for large GORs, then GOR gradients become large, and the solubility term contributes strongly to create asphaltene gradients.[35] There is a simple explanation why, for equilibrated oil columns, low GOR yields low GOR gradients while high GOR yields large GOR gradients.[2,3] The issue comes down in large part to compressibility. For compressible oils such as those with high GOR, the hydrostatic head pressure of the oil column creates a density gradient in the

2 SPE 146649

fluid column. This density gradient then creates a chemical compositional gradient forcing low density species such as methane to the top of the column. In contrast, low GOR black oils and heavy oils have very low compressibility, thus the hydrostatic head pressure does not create a density gradient. The lack of a density gradient keeps the compositional gradient small; thus, for this case, methane tends to be uniformly distributed. The gravity segregation of colloidal asphaltene is a fundamentally different type of effect and is related simply to Archimedes buoyancy. Archimedes buoyancy for methane molecules is small compared to asphaltenes. Archimedes buoyancy depends on species molar volume (cf. Eq. 1); asphaltene nanoaggregates have ~300 carbon atoms, methane only 1. DFA Field Studies of Connectivity via the FHZ EoS: Condensates. For condensates, asphaltene concentrations are low and thus asphaltenes are dispersed as molecules; the small size means that the gravity term is responsible for only a small part of the asphaltene gradient. However, the large GOR of condensates is consistent with large GOR gradients, meaning that the solubility term is responsible for a large asphaltene gradient. A condensate field was intersected by two wells and a side track. The primary operator concern was reservoir connectivity. The asphaltene concentrations in the condensate were measured at different heights in three wells using DFA; the asphaltenes found to be distributed across the field according to a molecular dispersion of asphaltenes in the crude oil and in accordance with the FHZ equation. This is expected for a condensate. For our purposes, it is not important whether the condensate heavy ends are the lightest asphaltenes or the heaviest resins. The chemical difference in these two categories is largely definitional.[36] A salient question for application of the FHZ EoS is whether the dispersion is molecular or colloidal. For condensates, we find that the colored heavy ends are molecularly dispersed.[33]

3660

3665

3670

3675

3680

3685

3690

3695

3700

3705

3710

De pt

h (T

VD -m

)

Optical Density at 647 nm

CFA Well A CFA Well B LFA Well A LFA Well B LFA Well C CFA Well C

O il Colum

Statoil

Nanoaggregate Cluster

Molecule0 10.5

Figure 10. The DFA-measured asphaltene distribution in three wells in the reservoir is largely

accounted for by the FHZ EoS (solid line).[33] The molecularly dispersed asphaltenes are equilibrated indicating the reservoir is connected. Subsequent production proved this prediction correct. Figure 10 shows DFA color (thus asphaltene) variations in three wells. The data is in accordance with the FHZ EoS for a molecular dispersion of asphaltenes with the exception of one DFA station at x682 meters.[33] The relatively large GOR gradient of the condensate [37] creates the large asphaltene gradient. The equilibration of the asphaltenes implies reservoir connectivity. This reservoir has two separate gas caps with two different GORs; the light ends are not equilibrated across the field. Thus, the cubic EoS does not make a clear prediction about reservoir connectivity. Since the asphaltenes do not partition to the gas phase, they remain unperturbed from the two different GOCs. Connectivity was proven in production. Connectivity: Black Oils. Black oils are often characterized by low GOR. In such a case the cubic EoS is not useful. Indeed the primary variation among black oils is asphaltene content. The FHZ EoS is well suited to handle black oils. Figure 11 shows DFA results from a field study of a reservoir intersected by many wells. The reservoir had two stacked sands that are not in pressure communication, thus not in flow communication. Inspection of Fig. 11 shows that the shallower sand has a slightly higher asphaltene content especially in conjunction with FHZ prediction, indicating compartmentalization corroborating the pressure data. The DFA color data from each sand show a gradient across almost the entire field that matches the FHZ equation; the asphaltenes are dispersed as nanoaggregates. Due to the low GOR, the gravity term dominates so an approximate analysis is very simple.[38] The ratio of asphaltene concentrations at two heights is just given by Archimedes buoyancy in the Boltzmann distribution: exp{-ΔρVgΔh/RT). Equilibrated asphaltenes indicate reservoir connectivity in each sand, which was subsequently confirmed in production. Many other case studies have exhibited nanoaggregate dispersion.[39,40]

0 2 3

TVD Feet 103

DFA Color & Asphtene Content 1

Nanoaggregate

2~3 nm

Cluster

4~6 nm

~1.5 nm

Molecule

O il Colum

FHZ EoS

FHZ

Figure 11. Black oil reservoir with two stacked

SPE 146649 9

sands. The asphaltene content across almost all the reservoir is accounted for by the FHZ EoS. The prediction of connectivity based on asphaltene equilibration throughout the reservoir was confirmed in production. Connectivity: Heavy Oil. Heavy oil has been a special challenge to any EoS due to the very high asphaltene content. Previously, there was not a clear method to handle the large gradients in heavy oil columns. However, with the nanoscience picture in Fig. 1 coupled with the FHZ EoS, the modeling of heavy oil gradients is straightforward. Heavy oils that still flow are accounted for by the Yen- Mullins model of colloidal structure. With increasing asphaltene content or colder temperatures, the heavy oil can become viscoelastic and will not flow. In this situation, larger length scale structures than clusters are present. Indeed, this rheological property is desired for pavement; at high temperatures the organic component of asphalt (with its high asphaltene content) can flow, thus can be applied, and at low temperature asphalt gels and becomes viscoelastic.[41]

2290

2295

2300

2305

2310 10 12 14 16 18 20

Asphaltene (wt%)

Nanoaggregate

2~3 nm

Cluster

5 nm

~1.5 nm

Molecule

TV D D ep

th ( m et er s)

O il Colum

Figure 12. Heavy oil column exhibiting a very large asphaltene gradient. Asphaltenes are dispersed in heavy oils as clusters due to the high asphaltene concentration. The clusters being large yield very large gravitational gradients of asphaltenes. The vertical equilibration of asphaltenes in this single well implies vertical connectivity, consistent with production data. The heavy oil in Fig. 12 gives a gradient 50 times larger than the low GOR black oil in Fig. 11. In both cases, the dominate term in the FHZ EoS is the gravity term; exp{- ΔρVgΔh/RT). The only difference is that in the black oil, the asphaltenes are present as nanoaggregates while in the heavy oil, the asphaltenes are present as clusters. Even though the diameter of the cluster is only about 2.5 times bigger than that of the nanoaggregate, the volume depends on the cube of this difference, and the volume

enters an exponential in the gravity term. This yields the huge (x50) mulplicative difference in asphaltene gradient. What was once considered to be a difficult problem – understanding gradients in heavy oil columns – is now seen to be founded on a very simple framework. Of course, disequilibrium can occur in oil columns including heavy oil columns. Nevertheless, the equilibrium case provided by the FHZ EoS gives a starting point for any oil column. Vertical and or lateral disequilibrium can then be analyzed (see below). Viscosity depends exponentially on asphaltene content.[5,41] The viscosity gradient for the heavy oil column in Fig. 11 is from 6 centipoise (cp) to 200 cp. Clearly, productivity predictions depend on an accurate assessment of asphaltene gradients particularly in heavy oil. Tar Mats. The term tar or “tar mat” is used inconsistently and confusingly in the oil industry and whereas the term tar commonly refers to pyrolysis product of organic materials, “tar mats” in oil and gas industry usually refer to zones of asphaltic phase or asphaltic oil in lighter oil reservoirs. Asphaltic or asphaltene rich oil in a reservoir can arise in many ways for example by phase instability and in-reservoir precipitation [42], or by severe biodegradation, or from primary oil charge at low maturity from very sulfur rich source rocks. The mechanism of formation of tar mats has long been a puzzle in the oil industry. The Yen-Mullins model provides a simple solution to this mechanism. Here, we define tar mats as a large discontinuous increase in asphaltene content from the oil leg to tar mat with tar mats typically containing ca 50% asphaltenes compared to much lower asphaltene content in their associated oil legs. With this definition, the development of tar mats likely corresponds to phase instability of asphaltene that we address with our model. The puzzle associated with this instability is as follows. If the instability of asphaltenes occurs at the top of the oil column, then how does the precipitated asphaltene move through the porous medium to get to the base of the oil column, the normal location of tar mats? If the asphaltene phase instability of asphaltenes occurs at the base of the oil column, then what process could give rise to this phase instability? With water at the base of the column, some have been tempted to claim a role for water in tar mat formation. Contrary to some suggestions in the literature, biodegradation is not a likely candidate to induce phase instability of asphaltenes. It is well known that microbes preferentially remove n-alkanes and other alkanes from crude oil.[43] This process is opposite to the standard laboratory process to induce phase instability of asphaltenes, the addition of n-alkanes. That is biodegradation should 1) increase the asphaltene content and 2) make asphaltenes less likely to undergo phase

2 SPE 146649

instability. There are no proper observations or examples of classical tar mats in biodegraded oilfields.[42] The solution to the puzzle of tar mat formation involves having two, not one stable colloidal sizes of asphaltenes. As we have seen, as the concentration of asphaltenes increases, nanoaggregates form clusters. This is well known in the laboratory.[6-8] The other similar circumstance that can lead to cluster formation is the reduction of solvancy of the liquid phase for asphaltenes. That is, the reduction of asphaltene stability can lead to the process going from nanoaggregates to clusters and subsequently to flocs. Most importantly, clusters are stably suspended in the liquid phase of the crude oil; they are not seeking a solid surface to settle upon. Figure 13 shows the result of a black oil column that was subjected to a late gas and condensate charge. Geochemical analysis established that this crude oil has been gas washed.[44] This light hydrocarbon addition destabilized the asphaltene thereby forming clusters. Destabilized asphaltene was confirmed in a flow assurance study of this crude oil. The clusters being larger settle towards the base of the column. Some bitumen deposition in core also confirmed asphaltene instability. In this case, the operator checked all associated risk factors for production, no significant problems were found and production proceeded, with the knowledge that large viscosity gradients characterize this oil column.

O il C

o lu m n

Nanoaggregate Cluster

Molecule

Figure 13. This is a black oil column that has both nanoaggregates and clusters. The clusters formed due to some asphaltene instability associated with a late, light hydrocarbon charge into the reservoir. The large clusters settled towards the base of the column increasing the asphaltene gradient. Excessive gas or condensate charge into the reservoir can destabilize the asphaltene causing tar mat formation.[42] This happened in a reservoir initially filled with crude oil.[45] So much gas charged into the reservoir that the

asphaltenes were destabilized to flocs after accumulating as clusters at the base of the oil column. A thin section of core at the base of the oil column is shown in Fig. 14. Most importantly, the tar mat in the core is seen to rest on cement. The base of the oil column is not in contact with water; this suggests that water had nothing to do with this tar mat formation.[45]

TAR

Figure 14. Thin section of a core at the base of the oil column where excessive gas charge caused a tar mat to form. The tar mat has formed on cement; water is not in contact with this oil at this point, this suggests the process to form this tar mat did not involve water.[45] Asphaltene instability was confirmed by analyzing the produced fluids, where asphaltene was found as a separated phase in production oil.[45] A late gas charge was confirmed by showing that methane isotope ratios varied, methane is not equilibrated with biogenic methane pooling updip.[45] In general, tar mats are found in fields which frequently exhibit asphaltene instability issues on production.[42] Fluid Disequilibrium. Equations of state presume fluid equilibration. Nevertheless, steady state or transient process acting on reservoir fluids can be incorporated into equations thereby accounting for disequilibrium in reservoir oil columns. In one deepwater field, there is a well known large asphaltene gradient that has been captured in a photograph of 24 dead oils samples.[46] In addition, this field has a large GOR gradient that is also not equilibrated, again as shown by the variable methane carbon isotope ratios throughout the field.[46] The gigantic asphaltene gradient that is evident visually had not previously been accounted for by any theory. A simple model presuming methane diffusion into the top of a black oil column, coupled with the FHZ theory has now accounted for this oil column and for the array of dead oils corresponding to this column.[47]

SPE 146649 11

Figure 15. The huge asphaltene gradient in a single oil column from a deepwater field is now matched by a simple theory presuming methane diffusion into the top of an oil column from a late gas charge and application of the Flory-Huggins-Zuo Equation.[47] The modeling quantitatively matches the measured GOR and asphaltene gradients.[47]

There are other mechanisms that can produce disequilibrium. For example, significant asphaltene and light (C6-C12) hydrocarbon gradients can also be caused by biodegradation processes in very heavy oils, even with low GOR gradients, whereby extremely low diffusive mixing rates in very viscous oils allow the preservation of large biodegradation induced compositional gradients.[48] Conclusions

Asphaltenes have long been known to be a flow assurance concern. However, for the much more important issues of reservoir and fluid complexities there simply had been no way to treat or utilize asphaltenes from any first principles approach. Asphaltene science had previously been characterized by order of magnitude uncertainties that precluded any first principles approach in development of corresponding theoretical formalisms.

Asphaltene nanoscience has undergone a renaissance with key results being confirmed by a plethora of methods. Asphaltene nanoscience has been codified in a simple, robust model known as the modified Yen model or the Yen-Mullins model. The advances subsumed in this model have enabled development of the simple, powerful Flory-Huggins-Zuo equation of state to treat asphaltene gradients in oil reservoirs. When combined with the new technology downhole fluid analysis (DFA), a myriad of reservoir complexities have come into the purview of treatment by first principles. Assessment of reservoir connectivity is dramatically improved as established in several field studies. Large variations in the magnitude of asphaltene gradients are now understood within a simple framework. The foundation for the mechanism of tar mat formation is now understood, and can be applied to many reservoirs. Important origins of large scale disequilibrium can be treated with simple transient concepts with corresponding equations treating many chemical and physical parameters of the reservoir fluids. As is always the case, linking new science and new technology leads to powerful new advances. In spite of the fact that the advances in asphaltene science and in DFA are relatively recent, a broad range of topics are already addressed. This bodes well for gaining future insights about the formidable complexities of reservoirs and their contained fluids.

References

[1] McCain WD, The Properties of Petroleum Fluids, PennWell Pub. Co. Tulsa OK, (1990) [3] Høier L, Whitson C, (2001), Compositional Grading, Theory and Practice, SPEREE, 525-535, (2001) [3] Mullins OC, The Physics of Reservoir Fluids; Discovery through Downhole Fluid Analysis, Schlumberger Press, Houston, Texas, (2008) [4] Larter SR, Aplin AC, Reservoir geochemistry: methods, applications and opportunities. Geological Society, London, Special Publications, Vol. 86, (1995) [5] Sirota EB, Lin, MY, The physical behavior of asphaltenes, Energy & Fuels, 21, 2809-2815, (2007) [6] Mullins OC, Sheu EY, Hammami A, Marshall AG, (Editors), Asphaltenes, Heavy Oils and Petroleomics, Springer, New York, (2007) [7] Mullins OC, The modified Yen model, Energy & Fuels, 24, 2179-2207, (2010) [8] Mullins OC, The Asphaltenes, Annual Reviews of Analytic. Chem., 4, 393-418, (2011) [9] Sabbah H, Morrow AL Pomerantz AE, Zare RN, Evidence for Island Structures as the Dominant Architecture of Asphaltenes, Energy & Fuels, 1597-1604, (2011) [10] Ruiz-Morales Y, Aromaticity in Pericondensed Cyclopenta-Fused PAHs determined by Density Functional Theory NICS and the Y-Rule—Implications in Oil Asphaltene Stability. Can. J. Chem. 634 87, 1280– 1295, (2009) [11] Groenzin H, Mullins OC, Molecular sizes of asphaltenes 707 from different origin. Energy Fuels, 14, 677, (2000) [12] Pomerantz AE, Hammond MR, Morrow AL, Mullins OC, Zare RN, Two step laser mass spectrometry of asphaltenes, J. Amer. Chem Soc., 130 (23), pp 7216– 7217, (2008) [13] Goncalves S, Castillo J, Fernandez A, Hung J, Absorbance and fluorescence spectroscopy on the aggregation of asphaltene toluene solutions. Fuel 83, 1823, (2004) [14] Andreatta G, Bostrom N, Mullins OC, High-Q Ultrasonic Determination of the Critical Nanoaggregate Concentration of Asphaltenes and the Critical Micelle Concentration of Standard Surfactants, Langmuir, 21, 2728, (2005)

2 SPE 146649

[15] Freed DE, Lisitza NV, Sen PN, Song YQ, A study of asphaltene nanoaggregation by NMR. Energy Fuels 23, 1189–1193, (2009) [16] Sheu EY, Long Y, Hamza H, Asphaltene self- association and precipitation in solvents; AC conductivity measurements. Chapter 10 in Ref. 6. [17] Zeng H, Song YQ, Johnson DL, Mullins OC, Critical nanoaggregate concentration of asphaltenes by low frequency conductivity, Energy Fuels, 23, 1201–1208, (2009) [18] Goual L, Sedghi M, Zeng H, Mullins OC, Electrical conductivity of n-alkane asphaltenes and its relation to aggregation and clustering. Fuel, in press [19] Mostowfi F, Indo K, Mullins OC, McFarlane R, Asphaltene Nanoaggregates and the Critical Nanoaggregate Concentration from Centrifugation, Energy & Fuels, 23, 1194–1200, (2009) [20] Barré L, Jestin, J, Morisset A, Palermo T, Simon S, Relation between nanoscale structure of asphaltene aggregates and their macroscopic solution properties. Oil Gas Sci. Technol. 64, 617–628, (2009) [21] Eyssautier J, Levitz P, Espinat D, Jestin J, Gummel J, Grillo I, Barré L, Insight into Asphaltene Nano-Aggregate Structure Inferred by Small Angle Neutron and X-Ray Scattering, J. Phys. Chem. B., 115, 6827–6837, (2011) [22] Barré L, Simon S, Palermo T, Solution properties of asphaltenes, Langmuir, 24, 3709-3717, (2008) [23] Yudin IK, Anisimov MA, Dynamic light scattering monitoring of asphaltene aggregation in crude oils and hydrocarbon solutions, Chapter 17 in Ref. 6. [24] Private communication, Professor William W. Mullins (deceased) [25] Andrews AB, McClelland A, Korkeila O, Krummel A, Mullins OC, Demidov A, Chen Z, Sum frequency generation studies of Langmuir films of complex surfactants and asphaltenes, Accepted Langmuir [26] Mullins OC, Rodgers RP, Weinheber P, Klein GC, Venkatramanan L, Andrews AB, Marshall AG, Oil Reservoir Characterization via Crude Oil Analysis by Downhole Fluid Analysis in Oil Wells with Visible-Near- Infrared Spectroscopy and by Laboratory Analysis with ESI FT-ICR Mass Spectroscopy, Energy & Fuels, 21, 256, (2007) [27] Ruiz-Morales Y, Wu X, Mullins OC, Electronic Absorption Edge of Crude Oils and Asphaltenes

Analyzed by Molecular Orbital Calculations with Optical Spectroscopy, Energy & Fuels, 21, 944, (2007) [28] Mullins OC, Ventura GT, Nelson RK, Betancourt SS, Raghuraman B, Reddy CM, Oil Reservoir Characterization by coupling Downhole Fluid Analysis with Laboratory 2D-GC Analysis of Crude Oils, Energy & Fuels, 22, 496-503, (2008) [29] Pfeiffer T, Reza Z, Schechter DS, McCain WD, Mullins OC, Determination of Fluid Composition Equilibrium under Consideration of Asphaltenes – a Substantially Superior Way to Assess Reservoir Connectivity than Formation Pressure Surveys, Denver Colorado, SPE ATCE, (2011) [30] Stainforth JG, “New Insights into Reservoir Filling and Mixing Processes” in Cubit JM, England WA, Larter S, (Eds.) Understanding Petroleum Reservoirs: toward and Integrated Reservoir Engineering and Geochemical Approach, Geological Society, London, Special Publication, (2004) [31] Buckley JS, Wang X, Creek JL, Solubility of the Least-Soluble Asphaltenes. Chapter 16 in Ref. 6. [32] Freed D, Mullins OC, Zuo JY, Asphaltene gradients in the presence of GOR gradients, Energy & Fuels, 24 (7), pp. 3942-3949, (2010) [33] Zuo JY, Freed D, Mullins OC, Zhang D, Gisolf A, Interpretation of DFA Color Gradients in Oil Columns Using the Flory-Huggins Solubility Model, SPE 130305, Int. Oil & Gas Conf. Beijing, China, June, (2010) [34] This Flory-Huggins-Zuo name for the EoS has been promulgated in many papers in spite of protests from Dr. Zuo. [35] Zuo JY, Elshahawi H, Dong C, Latifzai AS, Zhang D, Mullins OC, DFA Assessment of Connectivity for Active Gas Charging Reservoirs Using DFA Asphaltene Gradients, accepted, SPE #145448, ATCE, (2011) [36] Indo K, Ratulowski J, Dindoruk B, Gao J, Zuo JY, Mullins OC, Asphaltene Nanoaggregates Measured in a Live Crude Oil by Centrifugation, Energy & Fuels, 23, 4460–4469, (2009) [37] Dubost FX, Carnegie AJ, Mullins OC, Keefe MO, Betancourt SS, Zuo JY, Eriksen KO, Integration of In- Situ Fluid Measurements for Pressure Gradients Calculations, SPE 108494, Int. Oil Conf. Ex., Veracruz, Mexico, (2007) [38] Mullins OC, Betancourt SS, Cribbs ME, Creek JL, Andrews BA, Dubost FX, Venkataramanan L, The colloidal structure of crude oil and the structure of reservoirs, Energy & Fuels, 21, 2785-2794, (2007)

SPE 146649 13

[39] Betancourt SS, Ventura GT, Pomerantz AE, Viloria O, Dubost FX, Zuo JY, Monson G, Bustamante D, Purcell JM, Nelson RK, Rodgers RP, Reddy CM, Marshall AG, Mullins OC, Nanoaggregates of Asphaltenes in a Reservoir Crude Oil, Energy & Fuels, 23, 1178–1188, (2009) [40] Pomerantz AE, Ventura GT, McKenna AM, Cañas JA, Auman J, Koerner K, Curry D, Nelson RK, Reddy CM, Rodgers RP, Marshall AG, Peters KE, Mullins OC, Combining Biomarker and Bulk Compositional Gradient Analysis to Assess Reservoir Connectivity, Org. Geochem. 41 (8), pp. 812-821, (2010) [41] Lin MS, Lumsford KM, Glover CJ, Davison RR, Bullin JA, The effects of asphaltenes on the chemical and physical characteristics of asphalt, Ch. 5 in Asphaltenes, fundamentals and applications, Sheu EY, Mullins OC, Eds. Plenum Press, New York, (1998) [42] Wilhelms A, Larter SR, Origin of tar mats in petroleum reservoirs. Part II: formation mechanisms for tar mats, Mar. Petrol GeoL 11,442-456, (1994) [43] Peters KE, Walters CC, Moldowan JM, The Biomarker Guide, Cambridge University Press, Cambridge, UK, (2005) [44] Mullins OC, Freed DE, Zuo JY, Elshahawi H, Cribbs ME, Mishra VK, Gisolf A, Downhole Fluid Analysis coupled with Asphalene Nanoscience for Reservoir Evaluation, Presented in Perth, Australia, SPWLA, (2010) [45] Elshahawi H, Latifzai AS, Dong C, Zuo JY, Mullins OC, Understanding Reservoir Architecture Using Downhole Fluid Analysis and Asphaltene Science, Presented, Colorado Springs, SPWLA, Ann., Symp., (2011) [46] Elshahawi H, Dong C, Mullins OC, Hows M, Venkataramanan L, McKinney D, Flannery M, Hashem M, Integration of Geochemical, Mud Gas and Downhole Fluid Analyses for the Assessment of Compositional Grading - Case Studies, SPE 109684, ATCE, Anaheim, CA, (2007) [47] Zuo JY, Elshahawi H, Dong C, Latifzai AS, Zhang D, Mullins OC, DFA Assessment of Connectivity for Active Gas Charging Reservoirs Using DFA Asphaltene Gradients, SPE 145438, ATCE, Denver, Colorado, (2011) [48] Larter S, Adams J, Gates ID, Bennett B, Huang H, The origin, prediction and impact of oil viscosity heterogeneity on the production characteristics of tar sand and heavy oil reservoirs. Journal Of Canadian Petroleum Technology, 47(1), 52-61, (2008)