asphaltenes—problematic but rich in potential...aggregation and pressure on crude oil emulsion ......

22 Oilfield Review

Asphaltenes—Problematic but Rich in Potential

Kamran AkbarzadehAhmed HammamiAbdel KharratDan ZhangEdmonton, Alberta, Canada

Stephan AllensonNalco Energy Services L.P.Sugar Land, Texas, USA

Jefferson Creek Shah Kabir Chevron Energy Technology CompanyHouston, Texas

A. (Jamal) Jamaluddin Kuala Lumpur, Malaysia

Alan G. MarshallRyan P. RodgersNational High Magnetic Field LaboratoryFlorida State UniversityTallahassee, Florida, USA

Oliver C. MullinsHouston, Texas

Trond Solbakken Hydro Gulf of Mexico, LLCHouston, Texas

For help in preparation of this article, thanks to BallardAndrews, Soraya Betancourt, Denise Freed and MartinHürlimann, Cambridge, Massachusetts, USA; Myrt E. Cribbs,Chevron Energy Technology Company, Houston; Tara Davies,James Du and Ray Kennedy, Edmonton, Alberta, Canada;Peter Eichelberger, Adomite Chemicals, Nalco EnergyServices L.P., Sugar Land, Texas; Doris Gonzalez, Houston;Gregory Kubala, Sugar Land; Jeremiah M. Purcell, NationalHigh Magnetic Field Laboratory, Florida State University,Tallahassee, Florida; and Carlos Alberto Torres Nava, Maturin, Venezuela.For help in preparation of the front cover, thanks to Edo Boek,Cambridge, England.CHDT (Cased Hole Dynamics Tester), MDT (ModularFormation Dynamics Tester), Oilphase-DBR and Petrel aremarks of Schlumberger.

In the oil field, asphaltenes are best known for clogging wells, flowlines, surface

facilities and subsurface formations. Laboratory analysis and field intervention

help producers avoid or remediate asphaltene deposition. New science is finding

ways to use these enigmatic hydrocarbon compounds to better understand

reservoir architecture.

> Asphaltene precipitation and deposition. Changes in pressure, temperature, composition and shearrate may cause asphaltene precipitation and deposition. These changes may be induced by a varietyof processes, including primary depletion, injection of natural gas or carbon dioxide, acidizingtreatments and commingled production of incompatible fluids. Asphaltenes may build up at manyplaces along the production system, from inside the formation to pumps, tubing, wellheads, safetyvalves, flowlines and surface facilities.

Precipitated solidsin the separator

Solids in subsea flowline

Subseawellhead

Asphaltene deposition in thenear-wellbore region

Solids buildup in wellbore

59819schD6R1:Format Layout V7.2 8/23/07 1:53 PM Page 22

Summer 2007 23

Fluid-property variations that commonly occurduring the production of oil, such as changes inpressure, temperature and composition, canprecipitate asphaltenes. Small amounts ofasphaltenes may adhere to formation grains,pumps, tubulars, safety valves and flowlineswithout disrupting flow, but thick deposits canbring production to a halt.1 Optimizing produc -tion in this case requires knowing the oilcomposition and the conditions under which itsasphaltenes will remain in solution.

The mere presence of asphaltenes in a crudeoil does not portend asphaltene-related produc -tion problems. Heavy oils, those with the greatestasphaltene concentrations, are usually stableduring production, and do not promote wellclogging. Asphaltene-precipitation problems aremore common in lighter oils that contain minoramounts of asphaltenes in reservoirs that are atpressures well above bubblepoint.

While asphaltenes have practical uses, such asmaterial for road construction, waterproofing androofing, and as curing agents and corrosioninhibitors, they are usually considered a menacein the oil field. The potential for asphaltenes toclog systems continues downstream, and is aconcern for refiners, because asphaltenes aresignificant constituents of the heavy oils that areincreasingly entering refinery processing streams.Asphaltenes also play a role in the stability of oil-water emulsions, and in formation wettability.

This article first defines asphaltenes and thenfocuses on their behavior through case studies inupstream environments. We describe laboratoryand modeling techniques for predicting theconditions under which asphaltenes will precipi -tate during production. Then we examine remedi -ation techniques applied when deposition cannotbe avoided. We also discuss how asphaltenes arebeing used to understand reservoir architecture.

What Are Asphaltenes?Asphaltenes are a class of components ofhydrocarbons.2 Naturally occurring hydrocarbonfluids are compounds that span a continuum ofcomposition from dry natural gas to tar. Acrossthat range, density and viscosity increasedramatically, and color changes from clear todeep brown as asphaltene content increasesfrom 0 to nearly 20%.

Certain properties of asphaltenes have beenknown since before the first commercial oil wellswere drilled. The term originated in 1837 whenJ.B. Boussingault defined asphaltenes as theresidue of the distillation of bitumen: insoluble inalcohol and soluble in turpentine.3 The definitionin use today is similar: insoluble in n-alkanes, such

as n-pentane or n-heptane, and soluble in toluene.Asphaltenes obtained in this way are dark-colored, friable solids with a density of about1.2 g/cm3. They are also infusible, meaning theyhave no defined melting point, but decomposewhen heated, leaving a carbonaceous residue.

Because asphaltene content is an importantfactor in determining the processing and refiningpaths of a crude oil, a convenient laboratorymethod has been developed to quantify theasphaltene fraction. This technique separatesdead oil, or oil that has lost its gaseouscomponents, into saturates, aromatics, resinsand asphaltenes (SARA) depending on theirsolubility and polarity (above).

> Separating crude oil into saturates, aromatics, resins and asphaltenes(SARA). In SARA fractionation (top), asphaltenes are separated from theother hydrocarbon components by adding an n-alkane such as n-heptaneor propane. The remaining components, called maltenes, are then furtherfractionated by passing the mixture through a column. Each component isremoved from the column by flushing with various solvents. Saturatedhydrocarbons, or saturates, are removed by flushing with n-alkane.Saturated means the molecule contains the maximum number of hydrogenatoms possible, with no double or triple bonds between the carbon andhydrogen atoms. Saturates are also called alkanes. The simplest suchmolecule is methane, [CH4]. Aromatics incorporate one or more rings of sixcarbon atoms and six hydrogen atoms. The simplest aromatic is benzene[C6H6]. Resins are a solubility class, and somewhat similar to asphaltenes.They are the nonvolatile polar component of crude oil that is soluble in n-alkanes and insoluble in liquid propane.

ResinsSaturates Aromatics

Maltenes

Crude oil

Asphaltenes

Adsorb on silica,elute with

Dilute with n-alkane

n-alkane Toluene Toluene/methanol

Precipitate

1. Amin A, Riding M, Shepler R, Smedstad E andRatulowski J: “Subsea Development from Pore toProcess,” Oilfield Review 17, no. 1 (Spring 2005): 4–17.

2. Mullins OC, Sheu EY, Hammami A and Marshall AG (eds):Asphaltenes, Heavy Oils and Petroelomics. New YorkCity: Springer, 2007.Mullins OC and Sheu EY (eds): Structures and Dynamicsof Asphaltenes. New York City: Plenum, 1998.Chilingarian GV and Yen TF: Bitumens, Asphalts, and Tar Sands. New York City: Elsevier Scientific PublishingCo., 1978.

3. Boussingault JB : “Memoire sur la composition desbitumens,” Annales de Chimie et de Physique 64 (1837):141. Cited in Auflem IH: “Influence of AsphalteneAggregation and Pressure on Crude Oil EmulsionStability,” Doktor Ingeniør Thesis, Norwegian Universityof Science and Technology, Trondheim, June 2002.

59819schD6R1:Format Layout V7.2 8/23/07 1:53 PM 3

The advantage of the SARA method is that it isa simple procedure that can be performed inmany laboratories. However, SARA analysis alsohas several disadvantages that become apparentwhen it is used for purposes beyond its originalintent. First, dead oil lacks the gaseous compo -nents that are dissolved in live oils, and so theresults are not representative of how the oilwould act under reservoir conditions. In addition,laboratory methods vary greatly, and solubility ofasphaltenes varies with the type of n-alkane usedto precipitate them. This means that a single oilcould have two or more SARA results dependingon the precipitant used. Because of its simplicity,SARA analysis has become a widespread meansfor comparing oils, but often, because variationsin laboratory technique are not reported,comparisons between laboratories may not bevalid. Although the SARA method is a reasonablefirst step for categorizing dead crude oils, itprovides insufficient characterization for bothdownstream, refining needs and for upstreamconcerns, where live-oil properties are needed.

The definition of asphaltenes as a solubilityclass rather than as a chemical class has madethem more difficult to study than lightercomponents. The lighter components ofhydrocarbons—saturates and some aromatics—have concisely defined chemical structures(below). However, the heavier components,

asphaltenes and their related compounds, resins,have often been lumped together as residue anddeemed unworthy of or too challenging forfurther examination.

While their chemical structure has been slowto come to light, the average composition ofasphaltenes as a class is fairly well-known.Elemental analysis shows they are composed ofcarbon and hydrogen in an approximate 1 to 1.2ratio, compared with the 1 to 2 ratio for bulkalkanes. Unlike most hydrocarbon constituents,asphaltenes typically contain a few percent ofother atoms, called heteroatoms, such as sulfur,nitrogen, oxygen, vanadium and nickel. As far asasphaltene structure is concerned, experts agreethat some of the carbon and hydrogen atoms arebound in ring-like, aromatic groups, which alsocontain the heteroatoms. Alkane chains andcyclic alkanes contain the rest of the carbon andhydrogen atoms and are linked to the ring groups.Within this framework, asphaltenes exhibit arange of molecular weight and composition. Thiscompositional characterization is accepted bynearly all asphaltene specialists, but leaves ampleroom for debate about the structure or size ofindividual asphaltene molecules.

The extent to which these heavy hydrocarbonconstituents are less well-defined and under stoodthan light ones is partly a reflection of the greatereconomic value enjoyed by the lighter ends and

partly of the tractable experimental methodscommonly used for light-end analysis. Standardlaboratory methods such as gas chromato graphycan characterize components of the lighter,simpler hydrocarbon compounds with carbonnumbers less than about 36. Even large alkanesare amenable to specialized chroma tography.However, in the realm of the asphal tenes,standard methods are often not applicable, soextraordinary measures are required to extractaccurate information about component structure.

The list of techniques that have been used tostudy asphaltenes and other heavy fractionsencompasses mass spectrometry, electron micro -scopy, nuclear magnetic resonance, small-angleneutron and X-ray scattering, ultrasonicspectroscopy, dynamic light scattering, fluo -rescence correlation spectroscopy, fluores cencedepolarization, vapor-pressure osmometry andgel permeation chromatography. Because thesemethods investigate various aspects of asphal -tenes under different conditions, it is notsurprising that they have produced disparatemodels of asphaltene molecules.

Today, two main types of measurements—mass spectrometry and molecular diffusion—produce the most consistent evidence onasphaltene molecular weight and size (nextpage). Mass spectrometry induces a charge onthe molecule, accelerates the resulting ion in anelectromagnetic field, and measures the charge-to-mass ratio. Various types of mass spectrometryhave different ways of ionizing molecules andaccelerating ions.4 A key resource for asphaltenestudies using mass spectrometry is the NationalHigh Magnetic Field Laboratory at Florida StateUniversity in Tallahassee (see “Asphaltene MassSpectrometry,” page 26).

In molecular-diffusion measurements,various techniques, especially fluorescencetechniques, track the diffusion of individualmolecules.5 Large molecules diffuse slowly, andsmaller molecules diffuse more quickly.Estimates of molecular diameter are interpretedto infer molecular weight by comparison withmodel compounds.

In the 10 years since these techniques havebecome available, the concept of the asphaltenemolecule has undergone a transformation.Because the asphaltene solubility classificationcaptures a broad range of molecular structures,it is impossible to define a single molecularstructure and size. However, a picture isemerging that honors results from severalmeasurement types. This latest thinking puts theaverage molecular weight at about 750 g/molwithin a range of 300 to 1,400 g/mol. That is

24 Oilfield Review

> Molecular structure of some saturates and aromatics. Saturates includemethane, pentane and heptane. Benzene is the simplest aromatic.

H

C

H

H H

Methane

H

H

H

H H

HBenzene

PentaneH

C

H

H

H

C

H

H

C

C

CC

C

C

C

H

H

C

H

H

C

H

H

H

C

H

H

H

C

H

H

C

H

H

C

H

H

C

H

H

C

H

H

C

H

H

Heptane

(continued on page 28)


Summer 2007 25

4. Boduszynski MW: “Asphaltenes in Petroleum Asphalts:Composition and Formation,” in Bunger JW and Li NC(eds): Chemistry of Asphaltenes. Washington, DC:American Chemical Society (1981): 119–135.Rodgers RP and Marshall AG: “Petroleomics: AdvancedCharacterization of Petroleum-Derived Materials by Fourier-Transform Ion Cyclotron Resonance Mass Spectrometry (FT-ICR MS),” in Mullins et al,reference 2: 63–94.Merdrignac I, Desmazières B, Terrier P, Delobel A and Laprévote O: “Analysis of Raw and HydrotreatedAsphaltenes Using Off-Line and On-Line SEC/MSCoupling,” presented at the International Conference on

Heavy Organics Deposition, Los Cabos, Baja California,Mexico, November 14–19, 2004.Qian K, Edwards KE, Siskin M, Olmstead WN, Mennito AS,Dechert GJ and Hoosain NE: “Desorption and Ionizationof Heavy Petroleum Molecules and Measurement ofMolecular Weight Distributions,” Energy & Fuels 21, no. 2 (March 2007): 1042–1047.Hortal AR, Martínez-Haya B, Lobato MD, Pedrosa JMand Lago S: “On the Determination of Molecular WeightDistributions of Asphaltenes and Their Aggregates inLaser Desorption Ionization Experiments,“ Journal ofMass Spectrometry 41, no. 7 (July 2006): 960–968.

5. Groenzin H and Mullins OC: “Molecular Size andStructure of Asphaltenes from Various Sources,” Energy & Fuels 14, no. 3 (May 2000): 677–684.Freed DM, Lisitza NV, Sen PN and Song Y-Q: “MolecularComposition and Dynamics of Oils from DiffusionMeasurements,” in Mullins et al, reference 2: 279–300.Andrews AB, Guerra RE, Mullins OC and Sen PN:“Diffusivity of Asphaltene Molecules by FluorescenceCorrelation Spectroscopy,” Journal of PhysicalChemistry A 110, no. 26 (July 6, 2006): 8093–8097.Wargadalam VJ, Norinaga K and Iino M: “Size andShape of a Coal Asphaltene Studied by Viscosity andDiffusion Coefficient Measurements,” Fuel 81, no. 11–12(July 2002): 1403–1407.

> Some of the techniques that yield consistent results for asphaltene molecular weight and size. Descriptions of each technique are simplifiedsummaries and are for illustrative purposes only. Mass spectrometry measurements (blue shading) give results in terms of molecular weight. Moleculardiffusion measurements (yellow shading) give results in terms of molecular size, or diameter. Some other techniques, such as vapor-pressure osmometryand gel permeation chromatography, which are successful in characterizing lighter hydrocarbon compounds, give inconsistent values for asphaltenemolecular weight.

Abundance

Arom

atic

ity

Mol wt.

AsphalteneFraction

Molecular Weight

Method name Brief descriptionMolecular weightor molecular size

ReferenceGraphic icon

Field-ionization massspectrometry (FI-MS)

Probe-vaporized asphaltenespassing through a high-energy electricfield are ionized. Their mass-to-chargeratio is used to generate a massspectrum.

800 g/mol Boduszynski, reference 4

Electrospray ionization, Fouriertransform ion cyclotronresonance mass spectrometry(ESI FT-ICR MS)

This Nobel Prize-winning ionizationmethod evaporates solvent fromsolute, allowing very large moleculesto go into the vapor phase.

Most between 400 and800 g/mol, with a range of 300to 1,400 g/mol.

Rodgers and Marshall, reference 4

Atmospheric pressurephotoionization massspectrometry (APPI MS)

Gas-nebulized samples are ionizedby light to measure the mass-to-charge ratio of the asphaltenes.

750 g/mol, with a range of400 to 1,200 g/mol

Merdrignac et al, reference 4

Field-desorption/field-ionization mass spectrometry(FD-FI MS)

An asphaltene sample deposited ona needle-like surface is desorbed andionized when heat and a high electricfield are applied. The ion mass-to-charge ratio is used to generate themass spectrum.

~ 1,000 g/mol with a broaddistribution

Qian et al, reference 4

Laser desorption ionization(LDI)

A laser pulse on a solid asphaltenesample creates a gas plume. Lowlaser power and low gas densitiesare required for accuracy.

800 to 1,000 g/mol Hortal et al, reference 4

Time-resolved fluorescencedepolarization (TRFD)

Rotational diffusion constants ofasphaltene molecules in solutionare measured by detecting therate of decay of induced polarization.

~ 2 nm diameter correspondingto 750 g/mol with a rangebetween 500 and 1,000 g/mol

Groenzin and Mullins,reference 5

Nuclear magnetic diffusion Asphaltene molecules diffuse in anNMR field. Diffusion time is relatedto molecular size.

~ 2.6 nm diameter. Somedimers, or pairs of molecules,yield the larger size.

Freed et al, reference 5

Fluorescence correlationspectroscopy (FCS)

Translational diffusion coefficients offluorescing molecules lead to a lengthscale corresponding to a sphere-equivalent hydrodynamic radius.

~ 2.4 nm diameter correspondingto 750 g/mol. Smaller forcoal asphaltenes.

Andrews et al, reference 5

Taylor dispersion diffusion The translational diffusion coefficientof molecules in laminar flow is relatedto molecular size.

~ 1.4 nm diameter for coalasphaltenes (same as Groenzinand Mullins, reference 5)

Wargadalam et al, reference 5


26 Oilfield Review

Remarkably, the mass corresponding to anyone molecular elemental composition, forexample, CcHhNnOoSs, is unique. This unique-ness holds for molecules up to 1,000 dalton(Da) in mass, and provided that the molecularmass can be measured to within about 100 parts per billion (ppb), or 0.001 of themass of a hydrogen atom.1 For example, twomolecules that differ in composition by C3 ver-sus SH4, both weighing 36 Da, differ in massby 0.0034 Da—less than seven times the massof an electron! However, this difference can beresolved by mass spectrometry.

Mass spectrometry can resolve and identifymolecules based on mass, provided that the molecule can be ionized, that is, charged, andvaporized. The electrospray ionization tech-nique, for which John Fenn won the 2002Nobel Prize, can either remove a proton froma neutral acidic molecule, M, to form an (M-H)–

ion or add a proton to a neutral basic moleculeto yield an (M+H)+ ion. Electrospray ioniza-tion can thus be used to access the polarmolecules in petroleum, such as asphaltenes.

The mass of the ions can be resolved withultrahigh accuracy by placing them in a

magnetic field. In a magnetic field, the ioncyclotron rotational frequency is inversely pro-portional to ion mass, and can be measured tothe required 100-ppb precision by Fouriertransform (FT) ion cyclotron resonance (ICR)mass spectrometry (MS). Introduced in 1974,FT-ICR MS offers 10 to 100 times higher massresolution and mass accuracy than other massanalyzers, and is the only mass-analysismethod capable of resolving the chemical con-stituents of petroleum.2

The FT-ICR mass spectrometers at theNational High Magnetic Field Laboratory(NHMFL) in Tallahassee, Florida, can resolve up to 20,000 different elemental compositions in a single mass spectrum. An example fromVenezuela shows the results of measurementson a heavy oil containing more than 17,000component species (below).

Asphaltene Mass Spectrometry

> Facilities and measurements at the National High Magnetic Field Laboratory (NHMFL) in Tallahassee, Florida. Instrumentation at the laboratory(left) consists of three large magnets, which are contained in horizontally oriented cylindrical housings and operate at magnetic-field intensities of9.4 teslas, 9.4 teslas and 14.5 teslas (from foreground to background). An example mass-spectrometry output (right) shows the number of negativelyand positively charged components that can be resolved from a South American heavy-oil sample.

200 300 400 500 600 700 800 900–900 –800 –700 –600 –500 –400 –300 –200

South American Heavy Crude Oil

Negative ions Positive ions

6,118 resolvedcomponents

11,127 resolvedcomponents

Mass/charge


Summer 2007 27

Once the species have been identified, thenext step is to sort them by heteroatom class,or numbers of N, O and S atoms. Moreover,every additional ring or double bond requiresthe loss of two hydrogen atoms. Thus, knowl-edge of the numbers of C and H atoms in amolecule determines its number of rings plus double bonds; this sum is known as the “double-bond equivalent,” or DBE, defining themolecular “type.” Once the DBE is known, theremaining carbons must be aliphatic, forexample, bound as CH2 or CH3.3

It thus becomes possible to characterize acrude oil by mapping its characteristics in

multiple dimensions, such as heteroatom rela-tive abundance, DBE and carbon number. Asan example of this technique applied tocrude-oil processing, the relative abundanceof species containing one sulfur atom can beplotted before and after hydrodesulfurization(above). These plots can be used to assess theeffectiveness of processing methods.

These high-resolution measurements facilitate new understanding of petroleum constituents. The new field of “petroleomics”is based on the premise that sufficiently com-plete knowledge of the chemical compositionof petroleum should enable correlation, and

ultimately prediction, of its properties andbehavior. The NHMFL group is laying thegroundwork for such correlations to charac-terize deposits, heavy ends and asphaltenes,changes introduced by distillation andhydrotreatment, corrosion and oil-water emul-sions.4 For example, the group has confirmedand extended prior independent optical meas-urements by Schlumberger researchers toshow that even the heavy-end petroleum frac-tion is composed primarily of molecules lessthan 1,000 Da in molecular weight.5 Themethod also enables interfacially active com-ponents to be determined so that wettabilityand its effects can be understood. Ongoingand future applications are aimed at theanalysis of downhole fluids to detect compart-mentalization and identify potentialproduction problems.

1. One dalton equals one atomic mass unit, and is defined as one-twelfth of the mass of an unboundatom of carbon-12.

2. Comisarow MB and Marshall AG: “Fourier TransformIon Cyclotron Resonance Spectroscopy,” ChemicalPhysics Letters 25 (1974): 282–283.Marshall AG, Hendrickson CL and Jackson GS:“Fourier Transform Ion Cyclotron Resonance MassSpectrometry: A Primer,” Mass Spectrometry Reviews17 (1998): 1–35.

3. Aliphatic compounds are organic compounds in whichcarbon atoms are joined in chains or rings, with no double bonds. Examples are n-pentane, n-hexane and benzene.

4. Marshall AG and Rodgers RP: “Petroleomics: The Next Grand Challenge for Chemical Analysis,” Accounts ofChemical Research 37, no. 1 (2004): 53–59.Rodgers RP, Schaub TM and Marshall AG:“Petroleomics: Mass Spectrometry Returns to ItsRoots,” Analytical Chemistry 77 (2005): 20A–27A.

5. Mullins et al, reference 2, main text.

> Multidimensional characterization of a crude oil. Once the elemental compositions of the componentsof a crude oil have been identified by mass spectrometry, the components are sorted by heteroatomclass. This example shows the characterization of those species containing one sulfur atom. Plottedon the vertical axis is the double-bond equivalent (DBE), which is related to the number of carbon andhydrogen atoms in a molecule. The plot on the left depicts the characterization of these sulfur-bearingmolecules in the crude oil, and the plot on the right displays the characterization after the oil hasundergone hydrodesulfurization.

DBE

30

20

10

0

40

Carbon number Carbon number

Asphaltene Feed Processed Asphaltene

20 30 40 50 60 20 30 40 50 60

Relative abundance, percent of total


compatible with a molecule containing seven oreight fused aromatic rings, and the rangeaccommodates molecules with four to ten rings(below). There is also evidence that someasphaltenes consist of multiple groups of ringslinked by alkane chains.6

Heteroatoms, which are largely contained inthe ring systems, can give the molecule polarity:the polarizability of the fused aromatic-ringsystems and the charge separation induced byheteroatoms cause the centers of neighboringasphaltene molecules to stick to each other,

while the outer chains are repulsed by the chainsof other molecules. Such a structure is consistentwith the Yen model suggested more than 40 yearsago, which also proposed stacking asphaltenefused-ring systems.7 However, the molecularweight of a single molecule is significantlysmaller—by a factor of ten—than the averageasphaltene molecular weight proposed in the1980s and 1990s. Only now is the Yen modelunderstood within a framework of asphaltenemolecular structure and aggregation.

Using the size and structure of an individualmolecule as a starting point, chemists can nowexplain how asphaltene molecules behave beforethey precipitate (next page). This aggregationbehavior depends on solvent type. Most labora -tory studies are conducted with asphaltenesdissolved in a solvent, such as toluene, but someare performed with asphaltenes in their nativecrude oil.

At extremely low concentrations, below 10-4 mass fraction, asphaltene molecules intoluene are dispersed as a true solution.8 Athigher concentrations in toluene, on the order of10-4 mass fraction, asphaltene molecules sticktogether to form nanoaggregates, or nanometer-sized particles. These nanoaggregates aredispersed in the fluid as a nanocolloid, meaningthe nanometer-sized asphaltene solids are stablysuspended in the continuous liquid phase. Theconcentration at which nanoaggregates form wasfirst revealed by ultrasonic methods, but hasbeen confirmed more recently by nuclearmagnetic resonance diffusion measurements andconductivity results.9 As concentration reachesapproximately 5 g/L, or mass fraction of 5·10-3,the nanoaggregates appear to form clusters.10

The clusters remain in stable colloidal suspen -sion until asphaltene concentration reaches amass fraction of roughly 10-2. At higherconcentrations in toluene, the asphalteneclusters flocculate, forming clumps, which are nolonger stable in toluene, and so they precipitate.

In crude oil, the picture is even less clear, forseveral reasons. Most experimental methodsbecome difficult to interpret in crude oils. Also,the presence of other compounds in crude oilaffects asphaltene solubility. A recent studypoints to evidence of asphaltenes as nanoaggre -gates in crude-oil samples analyzed downhole.11

Laboratory experiments, discussed in the nextsection, show how changes in pressure, tempera -ture or composition can cause asphaltenes incrude oil to flocculate and form thick deposits.However, in some crude oils, asphaltenes canremain in a stable “solution” at extremely highconcentration without precipi tating. Forexample, Athabasca bitumen can contain morethan 18% asphaltene in a stable viscoelasticnetwork.12 Experts agree that more work isrequired to characterize asphaltene behavior incrude oils.

Laboratory Precipitation Methods:Asphaltenes in Crude OilCrude oils that exhibit asphaltene precipitationand deposition during primary depletion aretypically undersaturated, meaning they exist in

28 Oilfield Review

> Asphaltene molecular structures. Shown here are three of the manypossible structures of asphaltenes, which are a class of moleculescomposed of grouped aromatic rings (blue) with alkane chains. Some ringsmay be nonaromatic. Many of the rings are fused, meaning the rings shareat least one side. Heteroatoms such as sulfur, nitrogen, oxygen, vanadiumand nickel may reside in the aromatic rings. The molecule on the leftcontains a heteroatom in the form of sulfur [S]. Some asphaltenes consist ofmultiple groups of rings linked by alkane chains. The molecule on the leftcontains two such groups—one with ten rings and one with a single ring.

S

Alkane chains

Asphaltene Molecules

> The asphaltene-precipitation envelope (APE) in pressure-temperaturespace. The asphaltene-precipitation envelope (red curve) delimits thestability zones for asphaltenes in solution. For given example reservoirconditions (red dot), primary depletion causes pressure to decrease. Whenpressure reaches the upper asphaltene-precipitation envelope, also knownas the asphaltene-precipitation onset pressure, the least-solubleasphaltenes will precipitate. As pressure continues to decrease, moreasphaltenes will precipitate, until the bubblepoint pressure is reached, andgas is released from solution. With continued pressure decrease, enoughgas has been removed from the system, and the crude oil may begin toredissolve asphaltenes at the lower asphaltene-precipitation envelope.(Modified from Jamaluddin et al, reference 16.)

Example reservoir conditionsLiquid

Increasing temperature

Incr

easi

ng p

ress

ure

Asphaltene-Precipitation Envelope

Liquid andasphaltenes

Liquid, vapor andasphaltenes

Liquid andvapor

Vapor-liquid equilibrium (bubblepoint)

Upper asphaltene envelope

Lower asphaltene envelope


Summer 2007 29

the reservoir at pressures higher than thebubblepoint pressure. These crude oils tend tohave low asphaltene content and high gascontent. During reservoir production at aconstant temperature, once pressure decreasesto intersect the asphaltene-precipitationenvelope (APE), also known as the asphaltene-precipitation onset pressure, dissolved asphal -tenes start to precipitate and potentially depositin the reservoir and flowlines (previous page,bottom). Typically, the amount of precipitatedasphaltene increases as the pressure decreases,and reaches a maximum at the bubblepointpressure. The pressure-temperature (P-T) line

6. Gray MR: “Consistency of Asphaltene ChemicalStructures with Pyrolysis and Coking Behavior,” Energy & Fuels 17, no. 6 (November 2003): 1566–1569.Peng P, Fu J, Sheng G, Morales-Izquierdo A, Lown EMand Strausz OP: “Ruthenium-Ions-Catalyzed Oxidation of an Immature Asphaltene: Structural Features andBiomarker Distribution,” Energy & Fuels 13, no. 2(March 1999): 266–277.

7. Yen TF, Erdman JG, and Pollack SS: “Investigation of theStructure of Petroleum Asphaltenes by X-Ray Diffraction,”Analytical Chemistry 33, no. 11 (1961): 1587–1594. Dickie JP and Yen TF: “Macrostructures of the AsphalticFractions by Various Instrumental Methods,” AnalyticalChemistry 39, no. 14 (1967): 1847–1852.

8. Schneider MH, Andrews BA, Mitra-Kirtley S andMullins OC: “Asphaltene Molecular Size by FluorescenceCorrelation Spectroscopy,” accepted for publication inEnergy & Fuels, 2007.

9. Andreatta G, Bostrom N and Mullins OC:” UltrasonicSpectroscopy of Asphaltene Aggregation,” in Mullins et al, reference 2: 231–258.

Sheu E, Long Y and Hamza H: “Asphaltene Self-Association and Precipitation in Solvents—ACConductivity Measurements,” in Mullins et al, reference2: 259–278Freed et al, reference 5.

10. Oh K and Deo MD: “Near Infrared Spectroscopy to StudyAsphaltene Aggregation in Solvents,” in Mullins et al,reference 2: 469–488.Yudin IK and Anisimov MA: “Dynamic Light ScatteringMonitoring of Asphaltene Aggregation in Crude Oils and Hydrocarbon Solutions,” in Mullins et al, reference 2: 439–468.

11. Mullins OC, Betancourt SS, Cribbs ME, Creek JL,Dubost FX, Andrews AB and Venkatarmanan L: “The Colloidal Structure of Crude Oil and the Structureof Oil Reservoirs,” accepted for publication in Energy &Fuels, 2007.

12. Yang X and Czarnecki J: “Tracing Sodium Naphthenatein Asphaltenes Precipitated from Athabasca Bitumen,”Energy & Fuels 19, no. 6 (November 2005): 2455–2459.

> A way of looking at aggregation behavior of asphaltenes with increasing concentration. Asphaltenes exhibit different aggregation properties dependingon whether they are dissolved in crude oil (purple) or toluene (yellow). Individual molecules are seen only at low concentrations, below 100 mg/L, or 10-4

mass fraction. As concentration increases, molecules stick together, first in pairs, then in greater numbers. Once concentration rises to about 100 mg/L, or10-4 mass fraction, the molecules form near-spherical nanoaggregates of eight to ten molecules stacked together. At higher concentration levels, greaterthan 5,000 mg/L, or mass fraction of 5·10-3, nanoaggregates form clusters, in which the bodies of the nanoaggregates do not overlap, but the alkane chainsof neighboring clusters may interact. These clusters may remain in stable colloidal suspension until concentration reaches a mass fraction of 10-2. Stabilitycan continue to even higher concentrations in crude oil, where clusters may form a viscoelastic network. However, in toluene, high concentrations causethe asphaltene clusters to flocculate.

1.5 2 to 4 100 300

Particle size, nm

Destabiliz

ed asphaltene su

spension

Asph

alte

ne m

ass

fract

ion,

or c

once

ntra

tion,

mg/

L

Individualmolecules

Nanoaggregates of8 to 10 molecules

Clusters ofnanoaggregates

Viscoelastic network Floccules, no longer in a stable colloid

Stabiliz

edas

phalt

ene su

spen

sion5.10-3, 5,000

10-4, 100

5.10-2, 50,000


delineating the precipitation conditions abovethe bubble point is called the upper boundary ofthe asphaltene-precipitation envelope.

As pressure continues to decrease below thebubblepoint pressure, solution gas is removedfrom the oil, causing the oil to become denserand more optically refractive. Depressurizationbelow the bubblepoint may lead to redissolutionof the previously precipitated asphaltenes if thesystem is vigorously mixed and if asphalteneredissolution kinetics are relatively fast.13 In thiscase, the P-T diagram features a lower boundaryof the asphaltene-precipitation envelope, belowwhich the asphaltenes redissolve into solution.However, because pressure-induced asphalteneredissolution kinetics can be slow, the lowerboundary of the asphaltene-precipitation envelopecan be difficult to identify experimentally.14

Identifying the conditions at which asphal -tenes precipitate is the first step in seeking asolution to a potential asphaltene problem. To

identify these conditions, two types of laboratoryexperimental programs are conducted in advanceof any field development plan. Initially, precipi -tation measurements are performed to determineasphaltene-precipita tion onset pressures atconstant temperature or onset temperatures atconstant pressure.15 Since the precipitation ofasphaltene does not necessarily lead toasphaltene sticking or deposition, it is importantto also conduct deposition measurements atrealistic production conditions of temperature,pressure, composition and shear. The depositiontests help to assess the deposition tendency ofpressure-induced asphaltenes and to estimatethe rate of deposition.

Various laboratory techniques have beendeveloped for studying asphaltene precipitationfrom live crude oil.16 Methods such as gravimetricprecipitation, acoustic resonance and filtrationhave been used to determine asphaltene-onsetpressure. Other techniques, such as light scat -

tering, high-pressure microscopy and particle-sizeanalysis, have gained wide acceptance within theflow-assurance community, and have becomeindustry standards for screening reservoir fluidsfor asphaltene, wax and hydrate stability. Eachtechnique measures a different property of thefluid as pressure decreases. Combining resultsfrom several methods increases confidence in thequantification of the APE.

In the gravimetric method, asphaltenesprecipitate and fall to the bottom of a pressure-volume-temperature (PVT) cell. At selectedpressure steps, samples of the remaining fluidare analyzed through SARA fractionation, andshow a decrease in concen tration of asphaltenes.The method provides data for an asphalteneconcentration-versus-pressure plot, withtransitions that correspond to the upper andlower boundaries of the asphaltene-precipitationenvelope. The accuracy of this method is limitedby the selection of pressure steps and theaccuracy of the asphaltene-concentrationmeasurements. Greater accuracy requires smallintervals between pressure measurements, sothe experiment may be time-consuming andrequires large volumes of reser voir fluid. Also,this method may be subjective in estimating theonset of asphaltene precipitation because theonset point may be missed if the pressure stepsare too large.

In one example, the gravimetric methoddetected asphaltene precipitation in an oil from the Middle East.17 Asphaltenes insoluble inn-pentane and asphaltenes insoluble in n-heptanewere precipitated by SARA fractionationremaining after the gravimetric method (aboveleft). Measurements were performed at thereservoir temperature of 116°C [240°F].

Another method, the acoustic-resonancetechnique (ART), measures changes in theacoustic properties of the fluid as asphaltenesdrop out of solution. The additional solids in themixture increase the stiffness of the system. Aspressure decreases, an acoustic receiver on oneend of a PVT cell detects acoustic resonancegenerated by an acoustic transducer at the otherend of the cell. The acoustic system is accurate to± 100 psi [0.69 MPa] and requires only 10 mL ofsingle-phase reservoir fluid. The ART is less time-consuming than the gravimetric method. As fordisadvantages, the resonance changes detectedby the ART are not unique to asphalteneprecipitation; the presence of other solids andvapor-liquid phase boundaries could causesimilar changes in acoustic properties. Also, thetechnique does not allow the fluid to be mixed,giving rise to potentially inaccurate onset

30 Oilfield Review

> Gravimetric detection of asphaltene precipitation in a Middle East oil.SARA fractionation determined asphaltene content of the fluid remainingafter precipitating asphaltenes using n-pentane (blue circles) and n-heptane(red squares). Both types of asphaltenes showed the same precipitationtendencies. The precipitation-onset pressure determined gravimetricallywas 42.75 MPa [6,200 psi] for both types. Asphaltene content in theremaining liquid continued to decrease until pressure reached 22.24 MPa[3,225 psi], corresponding to the bubblepoint pressure. Additional pressurereduction caused dissolved asphaltene concentration to rise until pressurereached the lower asphaltene boundary at 13.5 MPa [1,960 psi], after whichthe asphaltene content stabilized at its original level. (Modified fromJamaluddin et al, reference 16.)

0 10 20 30 40 50 60 700

0.5

1.0

1.5Lower asphaltene-onset pressure = 13.5 MPa

Bubblepoint pressure = 22 MPa

Upper asphaltene-onsetpressure = 43 MPa

Pressure, MPa

Asph

alte

ne, %

by

wei

ght

n-heptane insolubles n-pentane insolubles

Air bath

Gravimetric Method

Pressure greaterthan asphaltene-onset pressure

Pressure less thanasphaltene-

onset pressure


Summer 2007 31

measurements caused by heterogeneous distri -bution of asphaltenes. Furthermore, the methoddoes not detect the lower boundary of theasphaltene-precipitation envelope. This may bebecause the dissolution of asphaltene is agradual phase transition.

The acoustic-resonance technique was usedto examine asphaltene precipitation in the sameMiddle East oil (right). As with the gravimetricmethod, measurements were taken at thereservoir temperature of 116°C. The asphaltene-onset pressure obtained by the ART agrees withthe results obtained by the gravimetric method.

The light-scattering technique (LST), alsoknown as the DBR solids-detection system(SDS), uses near-infrared light to probe fluids asasphaltenes precipitate either isothermally withdecreasing pressure or isobarically withdecreasing temperature. In the Oilphase-DBRfluid sampling and analysis laboratory setup, thePVT cell used for this technique is a transparentglass tube containing a magnetically drivenmixer.18 A near-infrared (NIR) light source onone side of the cell generates light atwavelengths between 800 and 2,200 nm at aspecific transmittance power. When asphaltenesprecipitate, they scatter light, reducing thetransmittance power of the light detected byfiber-optic sensors on the other side of the cell.As with the acoustic-resonance technique, theadvantages of the light-scattering method arespeed of testing and the low volume of single-phase reservoir fluid required.

Results of the light-scattering techniqueapplied to isothermal depressurization of an oilfrom the Gulf of Mexico show a typical response(right). The drop in the power of transmittedlight at 36.54 MPa [5,300 psi] marks the upper

13. Hammami A, Phelps CH, Monger-McClure T andLittle TM: “Asphaltene Precipitation from Live Oils: An Experimental Investigation of Onset Conditions and Reversibility,” Energy & Fuels 14, no. 1 (January 2000): 14–18.

14. Hammami A and Ratulowski J: “Precipitation andDeposition of Asphaltenes in Production Systems: A Flow Assurance Overview,” in Mullins et al, reference 2: 617–655.

15. Hammami and Ratulowski, reference 14.16. Hammami et al, reference 13.

Karan K, Hammami A, Flannery M and Stankiewicz A:“Evaluation of Asphaltene Instability and a ChemicalControl During Production of Live Oils,” PetroleumScience and Technology 21, no. 3 and 4 (January 2003):629–645.Jamaluddin AKM, Creek J, Kabir CS, McFadden JD,D’Cruz D, Manakalathil J, Joshi N and Ross B:“Laboratory Techniques to Measure ThermodynamicAsphaltene Instability,” Journal of Canadian PetroleumTechnology 41, no. 7 (July 2002): 44–52.

17. Jamaluddin et al, reference 16.18. Hammami A and Raines MA: “Paraffin Deposition from

Crude Oils: Comparison of Laboratory Results with FieldData,” SPE Journal 4, no. 1 (March 1999): 9–18.

Acoustic transducer

Acoustic receiver

Acoustic-Resonance Technique

0.25

Nor

mal

ized

acou

stic

resp

onse

0

0.50

0.75

1.00

0 10 20 30 40 50 60 70

Pressure, MPa

Upper asphaltene-onsetpressure = 43 MPa

Bubblepointpressure = 23 MPa

< Detecting asphaltene precipitationusing the acoustic-resonance technique.Acoustic-resonance measurements on aMiddle East oil show a sharp change inacoustic response at 42.92 MPa [6,225 psi],corresponding to the upper boundary ofthe asphaltene-precipitation envelope.The change at 22.68 MPa [3,290 psi] is thebubblepoint pressure. These results agreewith those obtained using the gravimetricmethod on the same oil. (Modified fromJamaluddin et al, reference 16.)

< Asphaltene-precipitation measurementson an oil from the Gulf of Mexico, usingthe light-scattering technique. As pressuredecreases from a high of more than 90 MPa[13,055 psi], the light transmission powerincreases, because the less dense fluidallows more transmission of light. At apressure of 36.54 MPa, the light transmit-tance signal (blue) plunges, signalingonset of asphaltene precipitation and theupper boundary of the APE. When pressurefalls to 33.09 MPa [4,800 psi], the transmit-tance falls even farther, as large clustersand floccules of asphaltene scatter alllight. At a pressure of 29.37 MPa [4,260 psi],

light transmittance increasesas bubbles of gas are createdat the bubblepoint. Thisresponse is in contrast to thatof some oils, which exhibitdecreased transmittance withthe appearance of bubbles.With continued depressuriza-tion, light transmittance jumpsat 26 MPa, when asphaltenesstart to redissolve. This is thelower boundary of the APE.(Modified from Jamaluddin etal, reference 16.)

Ligh

t tra

nsm

ittan

ce p

ower

, mW

0.25

0

0.50

0.75

1.00

0 20 40 60 80 100

Pressure, MPa

Near-infrared response

Bubblepoint pressure = 29 MPa

Upper asphaltene-onset pressure = 37 MPa

Lower asphaltene-onsetpressure = 26 MPa

Pressure greaterthan asphaltene-onset pressure

Pressure less thanasphaltene-

onset pressure

Lighttransmitter

Lightreceiver

Near-Infrared Light Scattering


APE, and the rise at 26 MPa [3,770 psi] points tothe lower APE. In another case, the method wasused with isobaric temperature decrease todetect asphaltene precipitation in a SouthAmerican crude (left).

For comparison, the Gulf of Mexico oil wasstudied through filtration experiments in anOilphase-DBR laboratory. In filtration measure -ments, the same PVT cell used in the light-scattering test is charged with 60 mL of single-phase reservoir fluid. A magnetic mixer agitatesthe cell contents as they are depressurized atreservoir temperature. At selected pressures, asmall amount of fluid is extracted from the celland passed through a 0.45-μm filter whilemaintaining pressure and temperature. SARAanalysis of the compounds trapped by the filtertracks the change in asphaltene content aspressure decreases.

An advantage of the filtration technique isthat it quantifies the amount of precipitatedasphaltene. The technique can be used to definethe upper and lower asphaltene phaseboundaries. Another advantage of the filtrationtechnique is that the asphaltenes are physicallyextracted from the oil, and so may be furthercharacterized through mass spectrometry,molecular-diffusion studies or SARA analysis.However, unlike other methods, results dependon filter size. The filtration technique alsorequires more time than measuring acousticresonance or light scattering.

Another advance, the DBR high-pressuremicroscope (HPM), allows direct visual obser -vation of multiple phases present at elevatedpressure and temperature. This technique allowsmicroscopic visualization of the appearance ofasphaltene particles as pressure decreases.19 Inone example, high-pressure micrographs helpedassess the effectiveness of different precipitationinhibitors on a crude oil from South America(left). The micrographs show an increase in thesize of asphaltene particles in the untreated oilas pressure decreases. The addition of chemicalinhibitor A changes the aggregationcharacteristics of the asphaltenes: asphalteneparticles become apparent at the same onsetpressure as for the untreated oil, but theparticles are much smaller, and remain smallereven as pressure continues to decrease.Inhibitor B is more effective at preventingasphaltene precipitation than inhibitor A.

32 Oilfield Review

> Light-scattering assessment of asphaltene onset during an isobarictemperature decrease. In this South American crude oil, asphaltenes beganprecipitating when the temperature reached 170°F [76°C]. Measurementswere conducted at 4,000 psi [27.6 MPa]. Also shown is the cloud point—thetemperature at which wax solidifies. (Modified from Hammami andRatulowski, reference 14.)

Pow

er o

f tra

nsm

itted

ligh

t, nW

Temperature, °F

500.01

0.1

1.0

10

100

70 90 110 130 150 170 190 210 230

Asphaltene onset~ 170°F

Cloud point = 91°F

Wellheadtemperature = 100°F

> High-pressure microscope (HPM) photographs of treated and untreatedSouth American crude oil as pressure decreases. In the untreated oil (toprow), dark particles interpreted as precipitated asphaltenes appear at apressure of 24.1 MPa [3,500 psi], and they grow larger as pressurecontinues to decrease. After treatment with 50 ppm of asphaltene-precipitation inhibitor A (second row), some particulate matter can bedetected again at 24.1 MPa. However, the particles are smaller, and remainsmaller than those in the untreated oil as pressure decreases. This indicatesthat the inhibitor had some effect on asphaltene stability. When the oil istreated with 50 ppm of inhibitor B (third row), asphaltene particles appear ata lower pressure than in the previous cases, and so chemical B is a moreeffective precipitation inhibitor. Treatment with 200 ppm of inhibitor B(bottom row) depresses the asphaltene-precipitation onset pressure evenfurther. (Modified from Karan et al, reference 19.)

Untreated

50 ppm A

50 ppm B

200 ppm B

17.2 MPa 20.7 MPa 24.1 MPa 27.6 MPa Sample ID

Decreasing pressure


Summer 2007 33

The HPM images can also be used to validate results obtained from the light-scattering technique. A break in the light-transmittance curve signifies a change in thefluid. HPM micrographs taken at selectedpressures help confirm the onset pointdetermined by the LST (right).

Although HPM is a direct and usefultechnique, it provides only a qualitativeindication of particle size and number. Toquantify these parameters, DBR scientists havedeveloped proprietary particle-size analysis(PSA) imaging software to analyze HPMphotographs. The PSA software scans digitalHPM images as they are acquired and deliversquantitative information on relative abundanceand size of particles, morphology changes andonset conditions.

Particle-size analysis on another crude oilfrom South America demonstrates the output ofthe image-analysis software (below right). Light-scattering measurements on this untreated oilhad determined the asphaltene-precipitationonset pressure to be 5,500 psi [37.9 MPa]. ThePSA results indicated a jump in particle size andparticle count as pressure decreased to theprecipitation-onset value. The size and numbercontinued to increase as pressure dropped even further.

Laboratory Measurements of Asphaltene DepositionAlthough asphaltene precipitation is a necessarycondition for the formation of obstructions, it isnot a sufficient condition. After precipitation,asphaltene particles must deposit and stick to a surface before they can become a flow-assurance problem.

While asphaltene precipitation is mainly afunction of temperature, pressure, fluid composi -tion and particle concentration, asphaltenedeposition is a much more complex process anddepends in addition on flow shear rate, surfacetype and characteristics, particle size andparticle-surface interactions.

19. Karan K, Hammami A, Flannery M and Stankiewicz A:“Systematic Evaluation of Asphaltene Instability andControl During Production of Live Oils: A Flow AssuranceStudy,” presented at the American Institute of ChemicalEngineers Spring National Meeting, New Orleans,March 10–14, 2002.

> Light-transmittance trace acquired during isothermal depressurization of a South American crudeoil. High-pressure micrographs taken at selected pressures yield direct evidence of asphalteneprecipitation. The asphaltene-onset pressure is 7,500 psi [51.7 MPa] at reservoir temperature,corresponding to the initial gentle drop in the power of transmitted light and the first observedasphaltene particles. Below about 4,000 psi, the precipitated asphaltene particles start to agglomerate,leading to the dramatic drop in the power of transmitted light.

Ligh

t tra

nsm

ittan

cePressure, psi

3,000 5,000 7,000 9,000 11,000 13,0001,0000

50

100

150

200

250

300

Saturationpressure,2,050 psi

> Particle-size analysis (PSA) generated during discrete depressurization of a South American crudeoil. HPM images (right) were taken at pressures above, equal to and below the asphaltene-precipitationonset pressure determined by independent measurements. The images were analyzed for the numberand size of particles and plotted in histogram form (left). The number and size of particles increasestrongly at 5,500 psi, which is the asphaltene-onset pressure. As pressure decreases, particle sizeand number increase. The small but finite particle count detected by PSA at the pressures above the onset pressure is attributed to the presence of impurities, such as water droplets, in the oil. Theequivalent diameter is the diameter of a circle with an area equivalent to the area of the observedparticle, and the count number is a cumulative value obtained from analysis of 20 images taken at thesame pressure. (Modified from Karan et al, reference 19.)

P = 6,500 psi

P = 6,000 psi

P = 5,500 psi

P = 5,250 psi

P = 5,000 psi

P = 4,500 psi

P = 4,000 psi

P = 3,500 psi100

10,000

Parti

cle co

unts

Size, µm

10 10 20 30 40 50 60 70 80 90 100 110 120

100

10,000

1

1001,000

110

1001,000

110

100

10,000

1

100

10,000

1

100

10,000

1

100

10,000

1


To investigate the tendency of organic solidsto deposit under actual flow conditions,Oilphase-DBR scientists have built the organicsolids deposition and control (OSDC) device(above).20 The rotational movement of a spindleat the center of the device produces a fluidmovement that creates a flow regime similar topipe flow.

Unlike other deposition measurementtechniques, the OSDC uses a relatively smallvolume—150 cm3—of fluid and can operate atpressures up to 103.4 MPa [15,000 psi],temperatures up to 200°C [392°F] and Reynoldsnumbers up to 500,000.21 The device can simulateproduction conditions of temperature, pressure,composition, surface type and either Reynoldsnumber or wall shear stress. Through carefullydesigned and machined cylindrical inserts, theOSDC can also mimic surface roughness oftubulars. These key parameters can be accu -rately and independently controlled, allowingthe deposition tests to be conducted over a widerange of conditions.

The OSDC device is often used to investigatethe effect of chemical inhibitors on the depositiontendency of asphaltenes. Chemical treatment isone of the commonly adopted control options forthe remediation and prevention of asphaltenedeposition in production tubing. A typicalinhibitor-selection test involves injecting aspecified concentration of inhibitor into thereservoir fluid enclosed in the PVT cell andmeasuring the reduction of asphaltene-onsetpressure of the inhibitor-treated oil as the fluidmixture depressurizes isothermally.

In one Gulf of Mexico example, a chemicalinhibitor was tested for its effectiveness inpreventing asphaltene deposition at reservoirtemperature and at a pressure close to itssaturation pressure. The initial screeningperformed by the chemical supplier on the stocktank oil sample had suggested that 200 ppmwould inhibit the deposition of asphaltenes.OSDC tests on the treated and untreated oilindicated that asphaltenes would deposit on theOSDC wall whether the oil was chemicallytreated or not (next page, top). After completion

of each test, the solids deposited on the wallwere collected and analyzed to determine theirasphaltene content using hot n-heptane. Theanalytical results showed that the suggestedasphaltene inhibitor could not fully preventasphaltenes from depositing on the wall,although it reduced the deposition rate ofasphaltenes by approximately 40%.

The OSDC is the only commercially availabledevice for studying the effects of asphaltenedeposition in live oils and realistic flow regimes,allowing more accurate assessment of chemical-additive requirements. In one case, a customerused OSDC results to effect a fivefold reduction inchemical usage. By reducing the concen tration ofadditive from 1,000 ppm to 200 ppm, annualchemical costs were cut by US $2.5 million.

A key factor in obtaining an accurateassessment of asphaltene precipitation anddeposition in live oils is the quality of the fluidsample. For asphaltene studies, as for all fluid-analysis programs, it is vital that the sample berepresentative of reservoir fluid and be main -tained at reservoir conditions throughouttransport to the laboratory. Any sample-collec -tion technique that fails to retain the single-phase nature of a sample risks significant errorsin the subsequent analysis, especially forasphaltenes. Anecdotes abound of operatorssurprised by severe asphaltene problemsbecause fluid samples had indicated noasphaltene content. Too late, they realized thatfluid sample collection had taken place atpressure below bubblepoint, and asphaltenesthat were in the fluid precipitated and adheredto the formation, leaving the sampled fluid freeof asphaltenes.

Single-phase samples can be acquired using avariety of techniques at different times in the lifeof the field.22 Common examples are the single-phase reservoir sampler (SRS) deployed duringdrillstem testing and the wireline-conveyed MDTModular Formation Dynamics Tester run with a single-phase multisample chamber in openhole.23 Representative samples can also beacquired with the CHDT Cased Hole DynamicsTester in cased hole.24

A high-quality fluid sample is not only single-phase, but is also free of contamination.Contamination can cause large errors in labora -tory measurements. For example, miscible oil-base mud (OBM) contamination in crude oil canchange the measured asphaltene-precipi tationonset pressure. Increasing the amount ofcontamination decreases the measured onsetpressure. In one case, just 1% OBM contamina -tion by weight caused asphaltene-precipitation

34 Oilfield Review

> The organic solids deposition and control (OSDC) device. The apparatus simulates production flowunder realistic conditions of pressure, temperature and composition for the study of asphaltene-deposition tendency. The OSDC tool helps optimize chemical treatment for the prevention andremediation of asphaltene deposition.

Multipointthermocouples

Shear cell

Variable-speedDC motor

High-pressureflow-controlvalves

Mechanicalmountingstand


Summer 2007 35

onset pressure to decrease by 0.7 to 1.0 MPa [100 to 150 psi] (right).25 These results under-score the need for low-contamination samples.

Modeling Asphaltene BehaviorLaboratory experiments to determine theasphaltene-precipitation envelope are usuallyconducted at reservoir conditions and also at asmall selection of other temperatures andpressures. To better understand asphaltenebehavior at the full range of conditions throughwhich the fluid will pass on its way to the surface,models have been developed that attempt tohonor experimental observations on asphalteneprecipitation and deposition. These models canbe divided into four groups: solubility models,solid models, colloidal models and associationequation-of-state (EOS) models.

Solubility models—Solubility models are theones most commonly applied to predictasphaltene precipitation. The first such model,established in 1984, employs a thermodynamicapproach to describe asphaltene stability interms of reversible solution equilibrium.26 Thisversion was easy to implement, but did notreproduce experimentally observed behavior.Since then, several researchers have craftedimprovements, mainly on the calculation ofasphaltene solubility parameters and thecharacterization of heavy fractions in crude oil.In the original version, a vapor-liquid equilibrium(VLE) was calculated to determine the

20. Zougari, M, Jacobs S, Ratulowski J, Hammami A,Broze G, Flannery M, Stankiewicz A and Karan K: “Novel Organic Solids Deposition and Control Device forLive-Oils: Design and Applications,” Energy & Fuels 20,no. 4 (July 2006): 1656–1663. Zougari M, Hammami A, Broze G and Fuex N: “Live OilsNovel Organic Solid Deposition and Control Device: WaxDeposition Validation,” paper SPE 93558, presented atthe SPE Middle East Oil and Gas Show and Conference,Bahrain, March 12–15, 2005.

21. Reynolds number is the ratio of inertial forces to viscous forces.

22. Aghar H, Carie M, Elshahawi H, Ricardo Gomez J,Saeedi J, Young C, Pinguet B, Swainson K, Takla E andTheuveny B: “The Expanding Scope of Well Testing,”Oilfield Review 19, no. 1 (Spring 2007): 44–59.

23. Jamaluddin AK, Ross B, Calder D, Brown J and Hashem M: “Single-Phase Bottomhole Sampling

Technology,” Journal of Canadian PetroleumnTechnology 41, no. 7 (2002): 25–30.

24. Burgess K, Fields T, Harrigan E, Golich GM,MacDougall T, Reeves R, Smith S, Thornsberry K,Ritchie B, Rivero R and Siegfried R: “Formation Testingand Sampling Through Casing,” Oilfield Review 14, no. 1(Spring 2002): 46–57.

25. Muhammad M, Joshi N, Creek J and McFadden J:“Effect of Oil Based Mud Contamination on Live FluidAsphaltene Precipitation Pressure,” presented at the 5thInternational Conference on Petroleum Phase Behaviourand Fouling, Banff, Alberta, Canada, June 13–17, 2004.

26. Hirschberg A, deJong LNJ, Shipper BA and Meijer JG:“Influence of Temperature and Pressure on AsphalteneFlocculation,” SPE Journal 24, no. 3 (June 1984): 283–293. Flory PJ: “Thermodynamics of High Polymer Solutions,”Journal of Chemical Physics 10, no. 1 (January 1942): 51–61.

> Assessing the effectiveness of an asphaltene-deposition inhibitor using OSDC tests. In one test, live oil from the Gulf of Mexico was left untreated (left),and subjected to reservoir flow conditions. In the second test, an asphaltene-deposition inhibitor chemical was mixed with the live oil (right) and themixture was subjected to the same flow conditions. The OSDC results indicated both untreated and treated oils would deposit asphaltenes. However, theasphaltene-deposition inhibitor reduced the deposition rate of asphaltenes to 7 mg/h, compared with 12 mg/h in the untreated case.

Untreated: 12 mg/h Treated: 7 mg/h

> The effect of sample contamination on asphaltene-precipitation onsetpressure. Measurements on hydrocarbon samples contaminated by oil-base drilling fluid can underestimate asphaltene-onset pressure. (Adapted from Muhammad et al, reference 25.)

Saturationpressure

Asphaltene-onsetpressures

Contamination,% by weight

(live-oil basis)2.67.614.219.4

Pressure

Pow

er o

f tra

nsm

itted

ligh

t


properties in liquid phase; then a liquid-liquidequilibrium (LLE) calculation—treating theasphaltene as a pseudoliquid—was performed,assuming no influence of the precipitatedasphaltene phase on the previously calculatedVLE. Later, researchers took the effect ofasphaltene precipitation on the gas phase intoaccount, and implemented a three-phaseequilibrium calculation.27 Work in 1995 extendedthe method to include polymer solutionthermodynamics.28 This more recent model canlead to a good representation of asphaltenebehavior if calibrated by experimental results,but may not accurately estimate asphalteneprecipitation in fluids with compositionsdifferent from those of the calibration crude oil.

Solid models—Solid models treat theprecipitating asphaltene as a single, solid-phasecomponent residing in a fluid whose phases aremodeled using a cubic EOS. Solid models mayrequire many empirical parameters and tuningto match experimental data. One solid modelassumes the crude oil’s heavy ends can bedivided into precipitating and nonprecipitatingcomponents.29 The precipitating components aretaken to be asphaltenes. This model is easy toimplement, but also requires experimental datato determine key parameters.

In another solid model, asphaltenes aretreated as a lumped pseudocomponent, and allother components are considered solvent.30 The

method is simple, and allows direct calculationof asphaltene solubility, but does not includepressure effects, which are especially importantto asphaltene stability.

Colloidal models—Colloidal models havetheir basis in statistical thermodynamics andcolloidal science. The first such model assumedasphaltenes exist in the oil as solid particles incolloidal suspension stabilized by resinsadsorbed on their surfaces.31 In this model, thevapor-liquid equilibrium calculated using an EOS establishes the composition of the liquidphase from which asphaltene may flocculate.Asphaltene-precipitation measurements at oneset of conditions are interpreted to give a criticalchemical potential for resins, which is subse -quently used to predict asphaltene precipitationat other conditions.

Association equation-of-state (EOS) models—The asphaltene precipitation model developed byDBR, now a Schlumberger company, is anassociation EOS model.32 This model makes fourmain assumptions: • Asphaltene molecules exist mainly as monomers

in the bulk crude oil and as aggregates in anassociated state in the precipitation phase;

• Asphaltene association leads to asphalteneprecipitation;

• The asphaltene precipitation process is thermo -dynamically reversible;

• The asphaltene-precipitation phase is a pseudo -liquid phase.

This model combines terms describing thechemical and physical effects of association ofasphaltene molecules. It requires composition,molecular weight, molecular size and interactionenergy of each component.

To date, most of the above asphaltene-precipitation models have been tested only onlimited sets of experimental results. Althoughmost authors claim that their model could givereasonable predictions, none can be used toconsistently predict asphaltene precipitation.However, these models all were developed beforethe emerging consensus on asphaltene molecularweight and structure. Until new models aredeveloped that incorporate the most recentexperimental findings, fluid chemists willcontinue to use the existing modeling methods.

Deepwater Gulf of Mexico Asphaltenes Recently in the Gulf of Mexico, Hydro Gulf ofMexico LLC encountered potential asphaltene-precipitation problems in a deepwater develop -ment. Reserves in the two reservoir intervals wereinsufficient to justify constructing two productionwells. A monobore subsea completion tappingboth layers would need to be tied back to anexisting platform for production to be economi -cally viable. Because commingling of fluidsresults in a composition change that could causeasphaltenes to fall out of solution, the fluids andthe entire production scenario had to be analyzedfor potential asphaltene problems. Furthercompositional changes could be induced byinjecting gas for gas lift, which was a completionmethod being considered. To reduce the risk offlow-assurance problems, scientists at Hydro,Schlumberger and Rice University in Houstonadopted a systematic approach for earlyevaluation of the potential impact of asphalteneprecipitation and deposition.33

Two wells penetrating the structure atdifferent depths encountered two differentfluids—oil and condensate. Laboratory measure -ments indicated that the two hydrocarbons werefrom different sources and were not in communi -cation with each other. SARA fractionation of theoil revealed relatively low asphaltene contentwith 61.7% saturates, 26.0% aromatics, 11.4%resins and 0.9% asphaltenes.

Laboratory measurements with the near-infrared light-scattering technique gave theasphaltene-onset pressure as 7,000 psi [48.3 MPa]±100 psi, and were corroborated by high-pressuremicroscopic photographs (left). However, the oilsample was known to contain 20% by weight OBM

36 Oilfield Review

> Near-infrared light-scattering measurements on a Gulf of Mexico oil from a Hydro well.Interpretation of the light-scattering readings (red dots) gave an asphaltene-onset pressure of 7,000 psi ± 100 psi. High-pressure microscopic photographs facilitated the visual interpretation ofasphaltene-particle appearance and corroborated asphaltene onset at a pressure between 7,500 and6,000 psi [51.7 and 41.4 MPa]. Bubbles were photographed at 3,000 psi, slightly below the bubblepointpressure determined by PVT analysis. (Modified from Gonzalez et al, reference 33.)

Ligh

t tra

nsm

ittan

ce

Pressure, psi

0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000

At 3,000 psiAt 6,000 psi

At 7,500 psi

At 9,500 psi

At 12,500 psi

Onset of asphalteneprecipitation atPres = 7,000 + 100 psi–Bubblepoint

pressure = 3,412 psi


Summer 2007 37

contamination. The onset pressure for anuncontaminated sample was estimated byextrapolation to be 10,700 psi [73.8 MPa].

Simulation helped engineers understand how the oil and its asphaltenes would behave throughout production—from theformation to the wellbore and then, oncecommingled with condensate, to the surface. Amolecular EOS model evaluated fluid atpressure, temperature and composition condi -tions along the 28,000-ft [8,534-m] length of theproduction system, and also predicted conditionsat key locations through the first five years of theproject (above).

Simulation results indicate that the black-oilreservoir fluid shows a high tendency toprecipitate asphaltenes as the pressure decreasesfrom its initial value of 16,988 psi [117 MPa](right). This tendency is exacerbated as gas

27. Kawanaka S, Park SJ and Mansoori GA: “OrganicDeposition from Reservoir Fluids: A ThermodynamicPredictive Technique,” SPE Reservoir EngineeringJournal (May 1991): 185–192.

28. Cimino R, Correra S, Sacomani P and Carniani C:“Thermodynamic Modelling for Prediction of AsphalteneDeposition in Live Oils,” paper SPE 28993, presented atthe SPE International Symposium on Oilfield Chemistry,San Antonio, Texas, February 14–17, 1995.

29. Nghiem L, Hassam M and Nutakki R: “Efficient Modellingof Asphaltene Precipitation,” paper SPE 26642,presented at the SPE Annual Technical Conference andExhibition, Houston, October 3–6, 1993.

30. Chung F, Sarathi P and Jones R: “Modelling ofAsphaltene and Wax Precipitation,” Topical Report inNIPER-498, DOE, January 1991.

33. Gonzalez D, Jamaluddin AKM, Solbakken T, Hirasaki Gand Chapman W: “Impact of Flow Assurance in theDevelopment of a Deepwater Project,” prepared for presentation at the SPE Annual TechnicalConference and Exhibition, Anaheim, California,November 11–14, 2007.

31. Leonartis KJ and Mansoori GA: “AsphalteneFlocculation During Oil Production and Processing: A Thermodynamic Colloidal Model,” paper SPE 16258,presented at the SPE International Symposium onOilfield Chemistry, San Antonio, February 4–6, 1987.

32. Du JL and Zhang D: “A Thermodynamic Model for thePrediction of Asphaltene Precipitation,” PetroleumScience and Technology 22, no. 7 & 8 (2004): 1023–1033.

> Modeling production conditions in the Hydro Gulf of Mexico commingling study. A schematic (left) shows the production system and nodes consideredin the integrated production modeling of the study. Temperatures, pressures and compositions along the entire production system—28,000 ft in length—were evaluated by a molecular equation-of-state (EOS) model. Pressures and temperatures were calculated for the reservoir, borehole (BH), comminglingzone, wellhead (WH) and separator over a five-year period (right). (Modified from Gonzalez et al, reference 33.)

180

Tem

pera

ture

, °F

Pres

sure

, psi

10,000

12,000

14,000

16,000

18,000

8,000

6,000

4,000

2,000

0

160

140

120

200

100

80

Pressure profile,early life

Pressure profileat 5 years

Temperature profiles,early life and at 5 years

GOR profiles,early life and at 5 years

Reservoir BH Commingle WH Separator

SeparatorFacilitiesRiser top

Riser

Riser bottomWellhead Flowline

Tubing

Commingle Internalcontrolvalve

Gas layer W

Gas layer ROil layer W

Oil layer R

Check-type valve

> Asphaltene-precipitation tendencies calculated for the Hydro comminglingstudy. The pressure-temperature trajectory (black) of the fluid as it isproduced from its initial temperature and pressure (red dot) intersects theasphaltene-onset pressure curve (solid blue) significantly above thebubblepoint pressure (red curve). This means the fluid has a high tendencyto precipitate asphaltenes during production, and will probably precipitatethem even if not commingled with fluids from the condensate reservoir.Experimental results on the oil sample, which was contaminated with oil-base mud (OBM), are shown as red triangles. Modeling calculated theasphaltene-instability curves for the OBM-contaminated sample (dottedblue curve) and for OBM-free oil. (Modified from Gonzalez et al, reference 33.)

Pres

sure

, psi

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

140 150 160 170 180 190 200 210

Temperature, °F

Asphaltene-instability

curves

OBM-free partial Pbubblepoint

OBM-contaminated partial Pbubblepoint

OBM-free

OBM-contaminatedExperimental data

Tres/Pres


condensate commingles with the black oil(above). The addition of condensate, whichincreases the gas/oil ratio (GOR) of the mixture,also increases the asphaltene-onset pressure.Without commin gling, the black oil willprecipitate approximately 10% of its asphaltenesunder isothermal depressurization to 10,000 psi[69 MPa]. At the commingling conditions, withlower temperature and higher GOR, asphalteneprecipitation increases to 60% at 10,000 psi.

Since the asphaltenes showed a propensityfor precipitation during primary depletion,injection of asphaltene-precipitation inhibitorswas advised. It was also recommended that the layers be produced sequentially: first the oil layer, then the condensate layer, with no commingling.

South Kuwait’s AsphaltenesA combination of sampling, laboratory analysesand modeling also helped Chevron understandasphaltene behavior in the Greater Burgan fieldof South Kuwait, where asphaltene deposition ontubulars and flowlines posed serious challengesto reservoir management and productionoperations.34 Fluid-analysis efforts centered ontwo objectives: characterizing the reservoir fluidto determine if asphaltene precipitation could beavoided, and evaluating solvents to mitigateproduction problems that could not be avoided.

For fluid characterization, four single-phasesamples were collected from four wells in theMarrat carbonate reservoir interval. Oilphase-DBR engineers conducted intensive analysis of

the sample from Well MG-OF4, and brief analysisof the samples from the other three wells. APIgravity of the Marrat oils varied between 36° and40°. SARA analysis on the live oil in Well MG-OF4indicated 68.3% saturates, 11.2% aromatics,18.4% resins and 2.1% asphaltenes. Results ofgravimetric measurements on MG-OF4 oildetermined asphaltene-precipitation pressure of6,200 psi [42.7 MPa] and bubblepoint pressure of3,235 psi [22.3 MPa] (next page).

Thermodynamic modeling treated theasphaltene as a solid phase in equilibrium withthe reservoir fluid. Simulating fluid behavior overthe range of conditions expected during produc -tion showed that precipitation of asphaltenefrom the Marrat oils could not be avoided, soengineers directed their search toward a cost-effective solvent for batch treatment or periodicinjection to minimize deposition on tubulars and flowlines.

38 Oilfield Review

> Changing asphaltene stability by commingling condensate with an asphaltene-prone oil. The addition of condensate (left) increases a black oil’s gas/oilratio (GOR), and increases the asphaltene-onset pressure. The asphaltene-onset curve of the original black oil (blue) is well below the oil reservoir’s initialpressure and temperature. As GOR increases with commingled production, both reservoirs fall below the asphaltene-onset curves, indicating the likelihoodof asphaltene precipitation in the wellbore. Another effect of increasing GOR is increased amount of asphaltene precipitation (right). A black oil with GORof 1,133 ft3/bbl [204 m3/m3] at reservoir conditions of 184°F [84°C] will precipitate 10% of its asphaltenes under depressurization to 10,000 psi (lower end ofred line). At the lower-temperature conditions of commingling at 170°F [77°C] and 10,000 psi, GOR increases to 1,364 ft3/bbl [276 m3/m3], and asphalteneprecipitation increases to 60%. (Modified from Gonzalez et al, reference 33.)

Pres

sure

, psi

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

150 160 170 180 190 200 210

Temperature, °F

Bubblepoint curves

GOR1,500 ft3/bbl

GOR1,250 ft3/bbl

GOR1,133 ft3/bbl

Gas and oilreservoir points

Asph

alte

ne p

reci

pita

ted,

mas

s fra

ctio

n

0.1

0

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1,000 1,100 1,200 1,300 1,400 1,500 1,600

GOR, ft3/bbl

Black-oilreservoir fluid

Commingle

Black oil andgas condensate

170°F

184°F

10,000 psi

> Using the acoustic-resonance technique (ART) to compare the ability ofvarious solvents to dissolve asphaltenes. The ART helped determineasphaltene-precipitation envelopes of five solutions. These are, in order ofincreasing ability to keep asphaltenes in solution: live oil (squares), live oilwith 20% deasphalted oil (triangles), live oil with 40% toluene (open circles),live oil with 40% deasphalted oil (diamonds) and live oil with 20%deasphalted oil and 1% dispersant (solid circles). (Modified from Kabir and Jamaluddin, reference 34.)

Pres

sure

, psi

7,000

3,000

4,000

5,000

6,000

8,000

9,000

150 175 200 225 250

Temperature, °F

LO

LO + 40% tolueneLO + 20% DAO

LO + 40% DAOLO + 20% DAO + 1% dispersantComputed wellbore temperature-pressure trace


Summer 2007 39

Because asphaltenes are by definition solublein toluene, toluene has been used successfully asan asphaltene solvent in well interventions.However, restrictions imposed by regulatoryorganizations now require operators to usesolvents that are more environmentally friendly.

As an alternative solvent, fluid analysts testedMarrat oil that had had its asphaltenes removed.For many oils, removal of asphaltenes byprecipitation is a reversible process. Onceasphaltenes have been removed by precipitation,the deasphalted oil has an increased capabilityto dissolve asphaltenes relative to the originallive oil.

The first step in assessing the usefulness of Marrat deasphalted oil in resolubilizing its ownasphaltenes was to remove the asphaltenes. Thiswas achieved by dissolving the live oil in n-pentane to precipitate the asphaltenes, col -lecting the filtrate and evaporating the n-pentane. SARA analysis of this deasphalted oil(DAO) yielded 59.5% saturates, 25% aromatics,15.3% resins and 0.2% asphaltenes—a reductionin asphaltene content of 90% compared with thelive oil.

Using the acoustic-resonance techniquedescribed earlier, engineers compared theasphaltene-onset pressures of the live oil withthose measured in solutions of live oil withvarying amounts of added DAO (previous page,bottom). The live oil with 20% DAO by volumedecreased the asphaltene-onset pressure by asmall amount, and an additional 20% DAOreduced the onset pressure even more. Thesolution with 40% DAO was more effective thanone with 40% toluene in decreasing asphaltene-onset pressure. However, the addition of 1%asphaltene dispersant improved the solvatingpower of the DAO even more significantly.

While the results indicated that deasphaltedMarrat oil might be able to dissolve Marratasphaltenes in the laboratory, applying thetechnique in the field proved challenging.Deasphalting large volumes of oil with n-pentanewas not feasible, so the procedure was modifiedto use condensate from a nearby gatheringfacility. The mixture was agitated and left atsurface conditions to allow its lightercomponents to evaporate. Oil deasphalted in thisway was nearly as asphaltene-free as thatdeasphalted by n-pentane, containing only 0.3%asphaltene by weight. However, reproducing the

method at the well location was difficult, and thesolution mixture pumped into the wellbore wasonly marginally deasphalted.

In spite of these difficulties, caliper measure -ments taken before and after treatment indicatedthat a substantial amount of asphaltene haddissolved from the borehole wall after a soakperiod of 24 hours. Not only was the treatmentmore environmentally friendly than other

methods, but it cost approximately 50% less thanusing toluene.

Unfortunately, success was short-lived,because reservoir pressure continued to decline,and asphaltene remediation was required moreoften. Treatment frequency increased from onceevery three months to once a month until thispart of the reservoir was shut down in 1998.

34. Kabir CS and Jamaluddin AKM: “AsphalteneCharacterization and Mitigation in South Kuwait’s Marrat Reservoir,” SPE Production & Facilities 17, no. 4(November 2002): 251–258.

> Gravimetric measurements on an oil from the Marrat formation of theGreater Burgan field in South Kuwait. The experiments began at pressureswell above the reservoir pressure of 7,700 psi [53 MPa]. As pressuredecreased to 6,200 psi, asphaltenes precipitated, indicated by the reductionin asphaltene content in the remaining fluid. The trend reversed at thebubblepoint pressure of 3,235 psi, when gas left the fluid. The concentrationof asphaltene at the lowest pressure is higher than in the original oilbecause asphaltenes are more soluble in the degassed oil. Asphaltenecontent in the remaining fluid is determined by SARA fractionation, whichcan use different alkanes to precipitate asphaltenes. Asphaltene contentprecipitated by n-heptane [C7H16] is shown as black squares, and asphaltenecontent precipitated by n-pentane [C5H12] is shown as red circles. (Modified from Kabir and Jamaluddin, reference 34.)

km

miles

60

60

0

0

Magwa, Ahmadiand Burgan fields

IRAQ

Kuwait City

KUWAIT

IRAN

SAUDI ARABIA

Raudhatain

Sabriyah

Minagish

Umm GudairAl Wafra

0.2

Asph

alte

ne, %

by

wei

ght

0

0.4

0.6

0.8

1.0

1.2

1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000

Pressure, psi

Bubblepointpressure = 3,235 psi

Asphaltene-onsetpressure = 6,200 psi

C5H12 Run 1C5H12 Run 2C7H16 Run 1C7H16 Run 2


Preventing Formation Damage by AsphaltenesAsphaltenes can deposit anywhere in the produc tion system, but perhaps the mostdamaging place is in the near-wellbore region,where asphaltene-blocked pores are difficult toaccess for remediation. Conventional asphalteneflocculation-inhibitor treatments involve eitherperiodic intervention with solvent soaks orcontinuous injection of chemicals into thewellbore. These methods are effective atpreventing agglomeration and deposition ofasphaltenes in flowlines and tubulars, but they donot protect the producing formation, because thechemicals interact with the oil after it has left theformation, potentially leaving asphaltenes behind.

An improved method developed by NalcoEnergy Services adds chemicals to the crude oilwhile it is still in the formation.35 The methodentails squeezing an asphaltene-depositioninhibitor into the formation to stabilize theasphaltenes before flocculation occurs. However,tests have shown that squeezing inhibitor alone

does not produce long-term benefits; formationsdo not absorb inhibitors adequately, allowinginhibitors to be quickly released from theformation as oil is produced. Pretreating theformation with an activator chemical enhancesabsorption of the inhibitor into the formationwithout changing formation wettability.

The general squeeze procedure includescleaning out and flowing back the well, pumpingin activator, a spacer of crude oil, inhibitor, andthen more crude oil, and shutting in the well for12 to 24 hours before resuming production(above).36 The activator prepares the formationand reacts with the inhibitor to make a complexthat remains in place for a prolonged period asthe well produces oil.

Nalco has applied this method and relatedtechnologies in areas with some of the mostserious asphaltene-deposition problems in theworld, including Venezuela, the Persian Gulf, theAdriatic Sea and the Gulf of Mexico. In oneexample from eastern Venezuela, severe

asphaltene-deposition problems caused a high-volume production well to plug within sevenmonths of treatment.37 Several cleaning methodshad been attempted, including physicallyscraping the wellbore and injecting xylene downthe tubing. Each cleaning event cost approxi -mately US $50,000 and two days of shut-inproduction. After squeeze treatment withactivator and inhibitor, the oil production rateincreased and the frequency of well cleaningdecreased to every eight months. The combina -tion of increased production and less frequentcleaning generated an annualized gain of60,882 barrels [9,674 m3], and a return on invest -ment of more than 3,000%.

Another case, this time from the Adriatic Sea,involved two deepwater subsea wells tying backto a floating production storage and offloading(FPSO) vessel. The operator learned of apotential asphaltene problem when the field wasinitially tested in 1993; examination of tubingstrings revealed thick deposits of asphaltenesalong a 3,300-ft [1,006-m] length of tubing,starting at a depth 6,500 ft [1,981 m] beneath theseafloor. Laboratory analysis of fluid samplesindicated that asphaltene deposition could becontrolled only by continuous downhole injectionof asphaltene dispersant. The appropriatetreatment program was designed and initiatedwith the desired results.

Once a successful treatment program wasunderway, additional laboratory work on samplescollected as part of a monitoring program helpedthe operator optimize dispersant dosage. Dose-response curves were generated to guide field-treatment procedures and give the best economicbalance between cost and the desired level ofasphaltene control. It was clear from surface-sample analysis that as dosage increased, thevolume of stable asphaltene dispersed in thecrude increased. This indicated that fewerasphaltenes were available to deposit in the

40 Oilfield Review

35. Allenson SJ and Walsh MA: “A Novel Way to TreatAsphaltene Deposition Problems Found in OilProduction,” paper SPE 37286, presented at the SPEInternational Symposium on Oilfield Chemistry, Houston,February 18–21, 1997.

36. Cenegy LM: “Survey of Successful World-WideAsphaltene Inhibitor Treatments in Oil ProductionFields,” paper SPE 71542, presented at the SPE AnnualTechnical Conference and Exhibition, New Orleans,September 30–October 3, 2001.

37. Cenegy, reference 36.38. Torres CA, Treint F, Alonso C, Milne A and Lecomte A:

“Asphaltenes Pipeline Cleanout: A Horizontal Challengefor Coiled Tubing,” paper SPE 93272, presented at theSPE/ICoTA Coiled Tubing Conference and Exhibition, The Woodlands, Texas, April 12–13, 2005.

39. Del Carmen Garcia M, Henriquez M and Orta J:“Asphaltene Deposition Prediction and Control in aVenezuelan North Monagas Oil Field,” paper SPE 80262,presented at the SPE International Symposium onOilfield Chemistry, Houston, February 5–7, 2003.

> An optimized squeeze procedure for treating a formation with activator and asphaltene-depositioninhibitor. The first step entails cleaning out and flowing back the well, then pumping in activator andan oil spacer. The activator binds to the formation. In the second step, the precipitation-inhibitorchemical is injected. The third step comprises a postflush with crude oil, and in the final step the wellis shut in for 12 to 24 hours, giving the activator and inhibitor time to form a complex before productionbegins. This method increases the residence time of the inhibitor in the formation. (Modified fromCenegy, reference 36.)

Step 4: Produce Well after Shut-In Period

Wellhead

Oil flowbackAsphaltene inhibitorreacts with activatorto form complex

Step 3: Squeeze Postflush of Oil

Wellhead

OilAsphalteneinhibitor

Oil spacer

Activator

Step 2: Squeeze Asphaltene Inhibitor

Wellhead

Asphalteneinhibitor

Oil spacer

Activator

Wellhead

Step 1: Squeeze Activator and Oil Spacer

Oil spacer

Activator

Channels information


Summer 2007 41

well (above left). A treatment level that optimizedcost and sufficiently stabilized the asphalteneswas shown to provide a protection level that was98% to 100% effective. Continuous treatment atthis level has enabled the wells to operate forseveral years without any plugging problems.

Mechanical Removal of AsphaltenesIn a field in the northern Monagas province ofeastern Venezuela, a combination of crude-oilcomposition and production conditions led tosevere pipeline clogging by asphaltenes (aboveright).38 During pipeline treatment, wellproduction was temporarily diverted to a mobiletesting unit, and the produced oil wastransported by truck.

Flow testing determined that two pipelinesections totaling 9,300 m [30,513 ft] in lengthwere completely plugged. Various cleaningoptions were considered, including high-pressurewater blasting, steam and xylene injection, andpipeline pigging units. All were eliminated fortechnical, environmental and economic reasons.The other alternative, replacing the pipeline,would cost US $1.4 million and take eight months.

A team comprising specialists from theoperating company and Schlumberger deter -mined that 2-in. coiled tubing (CT) couldpotentially clear the 85⁄8-in. outside diameter(OD) pipeline. An inclined injector-head frameallowed injection of the CT into the horizontallypositioned pipeline. The coiled tubing enteredthe pipeline from five different entry points.Water and water-base gel were pumped to carrythe dislodged solids, which came out in massive

quantities (above). By using CT instead of otheroptions to clean the pipeline, the client saved US $1 million and was able to return to normaloperations more quickly.

The asphaltenes that returned to surface didnot resemble asphaltenes typically precipitatedin laboratory tests. Analysis of solid organicdeposits in other fields in the region has shownthat the solids are not pure asphaltene, butcontain large amounts of other fractions. SARA

fractionation of eight samples in one northernMonagas field averaged 16% saturates, 15%aromatics, 25% resins and 44% asphaltenes.39

Making Use of AsphaltenesAnyone who deals with asphaltenes in the oilfield probably believes that the only positivevalue of asphaltenes in solution is that they havenot yet formed an obstructive deposit. However,asphaltenes, like many other hydrocarbon

> Optimizing asphaltene-dispersant dosage inthe Adriatic Sea. The volume of asphaltenedeposition decreased as dispersant dosageincreased. However, overtreating with dispersantincreases cost. Optimization requires acompromise that allows a tolerable amount ofdeposition at a reasonable cost. A treatmentlevel that allowed deposition of only 1% to 2% of the asphaltene volume enabled the wells tooperate for several years without asphaltene-deposition problems. (Modified from Cenegy,reference 36.)

Condition of CrudeAsphalteneDeposition

0 to 1% Crude strongly stabilized; dosagereduction indicated

1 to 2%

2 to 3.5%

> 3.5%

Crude well-stabilized; treatmentadequate; no dosage change indicated

Crude not perfectly stabilized; smallincrease in dosage indicated

Crude not stabilized; insufficientdosage

Pipeline

R i v e r

km

miles

1

1

0

0

> Use of coiled tubing to clear an 85⁄8-in. OD horizontal pipeline. Inclinationof the injector-head frame allowed the CT to be injected into the cloggedpipeline (right). Water and water-base gel helped to push out the organicsolids (left inset). (Modified from Torres et al, reference 38.)

> Extreme clogging of a surface pipeline in a field in theMonagas province of eastern Venezuela. A cut through asection of pipeline shows the severity of the cloggingproblem (left inset). Clearing the pipeline of asphaltenesrequired a technique that would be environmentallyacceptable, cost-effective and successful in the complexpipeline geometry (right). (Modified from Torres et al,reference 38.)


components, have the potential to revealimportant characteristics about the reservoir’sfluid, history and connectivity.

Chevron and Schlumberger used the opticalproperties of asphaltenes to understandreservoir connectivity in the Tahiti field, adeepwater structure in the Gulf of Mexico (left).The turbidite sands of the Tahiti field werediscovered in 2002 in 4,000 ft [1,219 m] of waterat true vertical depths ranging from 24,000 to27,000 ft [7,315 to 8,230 m].40 The reservoir layersdip steeply, having been tilted by salt tectonics.One appraisal well penetrated more than 1,000 ft [304 m] net of 600-mD pay. Fielddevelopment is currently projected to cost US $3.5 billion.41

The cost of any development depends on thenumber of wells required for optimal recovery,which in turn depends on the number of reser voircompartments. To check for reservoir compart -men tali zation, Chevron performed downholefluid analysis (DFA), which facilitates evaluationof fluid properties in real time. Downhole fluidanalysis helps identify compart mentalization byusing fluid-property signatures to determinewhether fluids are in communication.42

Three wells, each with at least one sidetrack,intersected the main Tahiti reservoirs atdifferent depths. Wireline optical spectroscopytools collected fluid samples and analyzedoptical density (OD) downhole at numerousdepths spanning a 3,000-ft [914-m] interval. Thehigh-quality OD measurements indicatedsignificant change in hydrocarbon colorationwith depth, implying a large variation in asphaltene content (left). Laboratoryassessment by SARA fractionation revealedasphaltene content increases from 1.6% byweight at the top of the reservoir to almost 6% at

42 Oilfield Review

New OrleansHouston

km

miles

150

150

0

0

Tahiti field

SandFine sandCoarse sandShale

N

> Downhole optical density (OD) varying with depth in the Tahiti field. Comparing downhole OD withlaboratory SARA-fractionation results produces a linear relationship between OD and asphaltenecontent (left). Optical density measurements in the two main sands, M21A (blue) and M21B (red),show a clear increase in OD with depth (right). Furthermore, the measured optical densities, whichare related to asphaltene concentrations, are consistent with solutions containing asphalteneparticles that have 1.3-nm and 1.5-nm diameters in M21A and M21B, respectively. (Modified fromMullins et al, reference 11.)

Asphaltene content by weight percent from laboratory SARA analysis0 1 2 3 4 5 6 7

Optic

al d

ensi

ty a

t 1,0

70 n

m

0

0.5

1.0

1.5

2.0

2.5

26,50026,000

2.0

1.5

1.0

27,00025,50025,00024,500

True vertical depth, ft

Diameter = 1.3 nm

Diameter = 1.5 nm

M21B

M21A

40. Carreras PE, Turner SE and Wilkinson GT: “Tahiti:Development Strategy Assessment Using Design ofExperiments and Response Surface Methods,” paperSPE 100656, presented at the SPE Western Regional/AAPG Pacific Section/GSA Cordilleran Section JointMeeting, Anchorage, May 8–10, 2006.

41. Baskin B: “Chevron Bets Big on Gulf Output,” The WallStreet Journal Online, June 27, 2007, page B5C,http://online.wsj.com/article/SB118291402301349620.html?mod=bolcrnews (accessed July 2, 2007).

42. Elshahawi H, Hashem M, Mullins OC and Fujisawa G:“The Missing Link—Identification of ReservoirCompartmentalization Through Downhole FluidAnalysis,” paper SPE 94709, presented at the SPEAnnual Technical Conference and Exhibition, Dallas,October 9–12, 2005.

43. Mullins OC, Betancourt SS, Cribbs ME, Creek JL,Dubost FX, Andrews AB and Venkataraman L:“Asphaltene Gravitational Gradient in a DeepwaterReservoir as Determined by Downhole Fluid Analysis,”paper SPE 106375, presented at the SPE InternationalSymposium on Oilfield Chemistry, Houston, February 28–March 2, 2007.Mullins et al, reference 11.

44. Betancourt SS, Dubost FX, Mullins OC, Cribbs ME,Creek JL and Mathews SG: “Predicting Downhole FluidAnalysis Logs to Investigate Reservoir Connectivity,”paper IPTC-11488-PP, to be presented at theInternational Petroleum Technology Conference, Dubai,UAE, December 4–6, 2007.

45. While the origin of this belief is not well-documented,many books and papers make this assumption.

46. Cimino R, Correra S, Del Bianco A, Lockhart TP:“Solubility and Phase Behavior of Asphaltenes inHydrocarbon Media,” in Mullins et al, reference 2: 97–103.Buckley JS, Wang J and Cree JL: “Solubility of theLeast-Soluble Asphaltenes,” in Mullins et al, reference 2:401–437.

47. Auflem IH: “Influence of Asphaltene Aggregation andPressure on Crude Oil Emulsion Stability,” DoktorIngeniør Thesis, Norwegian University of Science andTechnology, Trondheim, June 2002.

48. Kovscek AR, Wong H and Radke CJ: “A Pore-LevelScenario for the Development of Mixed Wettability in OilReservoirs,” American Institute of Chemical EngineersJournal 39, no. 6 (June 1993): 1072–1085.Yang and Czarnecki, reference 12.

49. “Molecular Management,” http://www.exxon.mobil.com/scitech/leaders/capabilities/mn_downstream_molecular.html (accessed July 30, 2007).Saeger RB, Quann RJ and Kennedy CR: “CompositionalModeling of Refinery Streams and Processes,”presented at the 232nd American Chemical SocietyNational Meeting, San Francisco, September 10–14, 2006.

50. Rodgers and Marshall, reference 4.

> The Tahiti field, a deepwater structure in the Gulf ofMexico. Chevron drilled three main wells, along withsubsidiary wells, to evaluate the extent and connectivityof the reservoir over a 3,000-ft depth interval. A modelcreated with Petrel seismic-to-simulation softwareshows sand and shale facies superimposed on thestructure of the top reservoir (left inset).


Summer 2007 43

the bottom. The linear relationship between ODand SARA-fractionated asphaltene contentproved that the variation in asphaltene contentis the main cause of OD variation.

The best explanation for the continuousgradation in asphaltene content is a single,continuous fluid compartment whose asphal -tenes have settled into equilibrium over geologictime.43 This high degree of reservoir connectivitywas good news for the operating company,because fewer wells will be required to developthis part of the field.

The OD measurements allowed reservoirspecialists to incorporate fluid properties into anexisting geological model over a large portion ofthe field.44 The resulting model was used topredict the fluid properties, including asphal -tene content, in a subsequently drilled well. Themeasured properties matched the predictions,confirming both the fluid model and reservoircontinuity. The asphaltene content observed inthe new well generally agreed with the largegradient seen in other wells.

A surprising result of further analysis of theTahiti fluids is that resin content appearsunrelated to asphaltene content (right).Asphaltene content indicated by OD measure -ments increases by more than 100%, while resincontent increases by only 8% from the top to thebottom of the reservoir. This finding runs counterto a long-standing and widely held tenet thatresins are associated with asphaltenes and arerequired for asphaltenes to be stable in crudeoil.45 Some asphaltene chemists have questionedthis tenet, knowing that asphaltenes can bestable in toluene without resins.46 Determiningthe role resins play, if any, in asphaltene stabilitymay help chemists develop better methods forpreventing and remediating asphaltene problems.

More Work on AsphaltenesAsphaltenes are best known for the problemsthey cause as solid deposits that obstruct flow inthe production system. However, asphaltenesalso cause other challenges to fluid flow: not onlydo they increase fluid viscosity and density, butthey also stabilize oil-water emulsions.47

Emulsions form when oil and water mix underconditions of agitation. Usually, the mixture ismore viscous than its components, and flows lesseasily. Separating emulsified water and oil isdifficult, and requires more than the gravita -tional methods used in most separators. A betterunderstanding of the effect of asphaltenes maybe the key to preventing the formation ofemulsions or tempering the deleterious effects ofthese mixtures.

Asphaltenes are also an important factor in determining formation wettability (see“Fundamentals of Wettability,” page 44). Changesin wettability can occur when even minisculeamounts of asphaltenes adsorb to formationgrains. In some wettability models, the presenceof asphaltene in the oil phase is required for thegeneration of oil-wetting conditions.48

Asphaltenes have the potential to derailupstream E&P activities, but they can also causedownstream disruptions, such as adhering to hotsurfaces in refineries. (For more on refining, see“Refining Review—A Look Behind the Fence,page 14.) A more comprehensive characterizationof asphaltenes and their properties is a priorityfor refiners, who hope to use molecular charac -teristics of asphaltenes and other hydrocarboncomponents in predictive compositional modelsfor refining and blending.49

Hydrocarbons are among the most complexfluids on earth. A single heavy-oil sample maycontain more than 20,000 chemical substances.The high resolving power and accuracy of newmeasurements, such as advanced massspectrometry, allow for the identification ofthousands of species in petroleum samples.

The way forward in asphaltene science relieson such techniques to better understand thestructure and function of these complicatedcompounds. The current state of asphaltene andcrude-oil characterization has been likened to astage in the evolution of protein science; proteinswere originally classified by solubility, but now,through the science of proteomics, theirfundamental structure, in terms of amino acids,is understood. Similarly, the term “petroleomics”has been coined for the study of the structure of hydrocarbons.50 The time will come when acrude oil will be characterized by all of itschemical constituents. —LS

> Resin and asphaltene optical signatures from shallow and deep crude-oil samples in the Tahiti field. In the original oils (bottom), where color isdominated by asphaltenes, the deep oil sample (red) has OD values that are twice those of the shallow sample (blue), indicating a large increase in asphaltene content with depth. After removing the asphaltenes byflocculation and precipitation, the remaining optical density can beattributed to resins. Resin content (top) shows little variation with depth in the reservoir. (Modified from Mullins et al, reference 11.)

2.0

1.5

1.0

0.5

0

Optic

al d

ensi

ty

Resins shallowResins deep

700 750 800400 450 500 550 600 650Wavelength, nm

3.0

2.5

2.0

1.5

1.0

0.5

0

Optic

al d

ensi

ty

Crude oil shallowCrude oil deep


asphaltenes—problematic but rich in potential...aggregation and pressure on crude oil emulsion ......

Documents