Journal of Petroleum Science and Engineering Volume 26, Issue 1-4, Pages 49–55, 2000

DOI: 10.1016/S0920-4105(00)00020-6Print ISSN: 09204105

Identification and Measurement of Petroleum PrecipitatesDynora Vazquez and G.Ali Mansoori

Thermodynamics Research Laboratory, UniÕersity of Illinois at Chicago, (M/C 063), Chicago, IL 60607-7052, USA

Abstract

A number of procedures to identify and measure the precipitates that result from petroleum fluids are presented. The

fractions considered in this study include those in the categories of asphaltenes, resins, saturates paraffinrwaxŽ ., aromatics, inorganic minerals and diamondoids. A combination of deposition techniques, separation by centrifuge, filtration, gaschromatography, gel-permeation chromatography, and SARA LC-HPLCŽ . separations, and a number of other techniques are utilized to identify each fraction and quantify their concentrations. These procedures provide an understanding of the overall behavior of the species that precipitate as well as of the interactions among them. The results of such analysis are the cornerstone of any predictive and preventive measures for heavy organics deposition from petroleum fluids.

Keywords: petroleum precipitates; deposition techniques; concentration

1. Introduction

Solid precipitation during production, transporta-tion, and storage of petroleum fluids is a commonproblem faced by the oil industry throughout theworld. Through complex phase transformations, dis-

Žsolved and suspended solids asphaltenes, resins,.paraffinrwax, diamondoids, formation solids, etc.

precipitate out of solution. Such phase segregationsare sometimes followed by flocculation of the result-ing precipitates. In many instances, these depositionphenomena render in complete clogging of flowlines and serious damages to storage vessels and

processing equipment. The solution or alleviation ofthe many technological problems posed by suchdepositions lies on a good understanding of themulti-phase behavior of the species that precipitate.It is also necessary to understand the interactionsamongst these various species in the local environ-ment where phase segregation and flocculation take

Ž .place Kawanaka et al., 1991; Mansoori, 1997 .The procedures presented in this report are com-

plimentary to the other procedures already in theliterature. The results of all such experimental mea-surements will be especially useful for the better

Žmodeling and prediction Kawanaka et al., 1991;.Mansoori, 1997 of the behavior of heavy organics in

petroleum fluids. In the following sections of thisreport, the procedures along with the results of acrude-oil analysis as performed by each procedureare presented.

Authors email addresses: D. Vazquez (dvazquezg@ipn.mx); G.A. Mansoori (mansoori@uic.edu)

50

2. Analysis of suspended materials in a crude oil

Microscopic inspection of crude oil samples isgenerally performed as a preliminary step. A micro-scope with a minimum magnification of 200= andequipped with a photo camera may be used. Pho-tographs are obtained by placing small samples ofthe crude oil in a micro-chamber that is then per-fectly sealed with a glass cover to prevent evapora-tion of the lighter fractions. This test will determinewhether a crude oil contains suspended solids. If not,measurements of the onsets of asphaltene precipita-

Žtion, flocculation and deposition Vuong, 1985; Es-.cobedo and Mansoori, 1995, 1997 and other related

measurements can safely be performed. Otherwise,such measurements cannot be performed unless thesuspended solids are found to be without asphaltene.It may be also necessary to establish the nature of

Ž .the droplets or bubbles present in the crude. Anumber of procedures may be devised to study these

Ždroplets through microscopic techniques Vazquez.and Mansoori, in press . These experiments may

demonstrate whether some of the species present incrude oil show surface activity at brine–oil interfaceindicative of stable water-in-oil emulsions in crudeoil. It must be noted that some of the species con-tained in a crude oil, especially asphaltene, mayexhibit surface activity that renders in the formationof stable water-in-oil emulsions. This emulsion seemsto be stabilized by a film that surrounds the brinedroplets. Generally suspended solids in the crude oilinteract with these droplets such that lumps of sev-eral droplets are held together by solid aggregates.The amounts of water and sediment content of acrude oil can be measured following the IP 359r82Ž .ASTM D4007-87 standard procedure using a cen-

Ž .trifuge IP 359r82, 1985 .For quantitative analyses, it is important to pre-

vent material loss due to evaporation during samplepreparation. It is necessary to devise an appropriateprocedure to draw aliquots from the main sample,such that uncertainties introduced by evaporationwill be minimized.

In order to ascertain reproducible results, allaliquots must be weighed in gas-tight vials.

The solids in a crude oil may be separated forŽfurther analysis by two different methods Vazquez

.and Mansoori, in press . In the first method, the

suspended solids are separated using an appropriatefiltration technique. In the second method, the solidsare separated using a centrifuge. Then, the amountsof nC -soluble, nC -asphaltene, and insoluble frac-5 5tions in the suspended solids are measured using a

Ž .detailed, procedure Vazquez and Mansoori, in press .Table 1 contains the results from the fractionation

Ž .of the suspended solids in a crude oil named A .Note from Table 1 that the suspended solids in crudeoil A are mainly comprised of nC -soluble material5that could very well be paraffinrwax in conjunctionwith resins and traces of oil trapped in the solid.Note also that these suspended solids contain a con-siderable amount of nC -asphaltenes as well as a5considerable amount of insoluble material. The com-position of the solid material contained in crude oilA suggests a strong interaction among insoluble

Ž .material minerals , asphaltenes, and paraffinrwax.The insoluble material contained in the suspendedsolids of crude oil A was further tested and found tobe inorganic minerals from the formation. Differen-tial scanning calorimetry is performed on the nC -5soluble material in the suspended solids in order toestablish its paraffinrwax molecular weight rangeŽ .Letoffe et al., 1995 . For crude oil A, the meltingcurve indicated that the maximum of the meltingoccurs in the range of 76.26–818C. These tempera-

Žtures correspond to nC hexatricontane; mps36. Ž .76.268C and nC tetracontane; mps81.008C .40

Thus, the thermogram suggested the suspended solidsin crude oil A contain mostly high-molecular-weightŽ .HMW paraffinrwax species.

It is not surprising that the suspended solids foundin crude oil A are primarily comprised of para-

Table 1Characterization of the suspended solids contained in crude oil A

Fraction Suspended solids Whole crude oil

Total amount 100 2.5909"0.1156 wt.%of solids

Ž .Total volume 100 3.82"0.06% vrvof solidsnC5 solubles 81.7669"1.7794 2.1198"0.1404 nC5Ž .wt.%Asphaltenes 10.9081"1.4704 0.2815"0.0253Ž .wt.%Insoluble solids 7.3250"0.3133 0.1892"0.0032Ž .wt.%

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000

51

ffinrwax since its cloud point is above room temper-Ž .ature ;338C . However, what is important here is

the fact that this solid material also contains a sub-stantial amount of precipitated asphaltenes. It is ap-parent from these results that mineral particles andparaffinrwax crystals assist the precipitation of as-phaltenes. A plausible scenario would be the exis-tence of attractive forces between mineral particlesand asphaltenes, which would result in the formationof larger particles on which paraffinrwaxes maydeposit.

3. Total asphaltene content of a crude oil

The exact value of the total asphaltene content ofa crude oil cannot be measured by any precipitationprocedure because of the fact that it will not com-pletely precipitate by the existing precipitation pro-cedures. However, the asphaltene that precipitatesusing different normal paraffin hydrocarbon solventscan be measured. Then, using such data in an appro-priate polydisperse deposition model for asphaltene,one may be able to estimate the true value of the

Žtotal asphaltene content of a crude oil Kawanaka et.al., 1991; Mansoori, 1997 .

Measurement of the asphaltene content of nC -in-5solubles, could be a challenging task. It is wellknown that some amount of resins will precipitatealong with the asphaltene fraction and insoluble ma-terial. It is also known that HMW solid paraffinrwaxwill precipitate out of solution once the asphaltenestarts to flocculate. The experimental procedure used

Žhere finds its basis in the IP143r90 ASTM D3279-. Ž .90 standard procedure IP 143r90, 1985 .The total nC -, nC -, and nC -asphaltene con-5 7 9

tents of crude oil A were measured according to thisprocedure. These results are summarized in Table 2.The total content of insoluble material in crude oil A

Table 2Total nC , nC , and nC asphaltene content of crude oil A5 7 9

Paraffin solvent wt.% Asphaltene

n-Pentane 1.0833"0.0050n-Heptane 0.5167"0.0059n-Nonane 0.3982"nra

Fig. 1. Normalized amounts of asphaltenes precipitated with vari-ous paraffin solvents. The values reported in this figure werenormalized with respect to the content of insoluble materialpresent in crude oil A.

is determined by the same procedure as for asphal-tene precipitation. Fig. 1 shows the normalized totalnC -, nC -, and nC -asphaltene content of crude oil5 7 9A along with four other crude oils.

( )4. Gel permeation chromatography GPC studiesof precipitated asphaltenes from a crude oil

Adsorbed resin material on the precipitated as-phaltene particles may lead to discrepancies in theMW determination. Therefore, a given precipitatedasphaltene fraction of a crude oil must be purified

Žbefore the measurement Vazquez and Mansoori, in.press .

For any particular method of determination, theobserved molecular weights suggest that asphaltenesform molecular aggregates, even in dilute solutions.This association is influenced by solvent polarity,asphaltene concentration and temperature. The mostwidely used method for determining the MW ofasphaltenes is GPC. It is understood, however, thatGPC provides relative molecular weights and de-pends entirely on the calibration standards usedŽ .Leontaritis and Mansoori, 1989 .

The molecular weight distribution of the nC -,5nC -, and nC -asphaltene fractions precipitated from7 9the crude oil A were obtained by a GPC proceduredesigned specifically for asphaltene. Each asphaltene

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000

52

Ž .fraction was dissolved in tetrahydrofuran THF at adefined concentration. A similar solution of the wholecrude oil A was also prepared. All samples wereequilibrated using continuous mixing for a pre-scribed duration. Care was taken to avoid contact ofthe samples with open air. It must also be pointedout that all samples were filtered through appropriatemembrane filters to remove any insoluble materialsfrom the solutions. Three injections of each solutionwere run through a GPC system. The mobile phasefor these analyses was THF at an appropriate flowrate. The mobile phase was degassed by filtrationand subsequent sonication with dry helium. Thetemperature for all the runs were kept constant andthe eluant was analyzed. The raw GPC data wereprocessed using a size exclusion chromatographysoftware. Calibration curves were obtained usingnarrow-dispersity polystyrene standards.

The molecular weight distributions of the nC -,5nC -, and nC -asphaltene fractions of the crude oil7 9A are reported in Fig. 2. Since asphaltenes aresolubility-class compounds, properties of the precipi-tated asphaltenes vary according to the solvent usedfor their precipitation. The resin content in the pre-

cipitate increases as the number of carbons in theprecipitating solvent decreases. When n-pentane isused HMW paraffins co-precipitate with the as-phaltenes and resins. Therefore, it is expected thatthese three types of asphaltenes will exhibit verydifferent MW distributions. The MW distributionswill depend on the level of interactions among as-phaltenes, resins, and co-precipitated HMW paraf-fins.

According to Fig. 2, the MW distributions for allthe three types of asphaltenes are bimodal. Thedistribution for the nC -asphaltenes exhibits two dis-5

w xtinguishable peaks: one narrow HMW peak 6751Ž .and one broad low-molecular-weight LMW peak

w x2663 . The HMW peak exhibits little tailing towardshigh molecular weights. The shape of these twopeaks suggests the existence of, at least, two differ-ent types of species in the solution.

The MW distribution for the nC -asphaltenes also7exhibits two distinct peaks: a narrow HMW peakŽ .9383 with considerable tailing towards the high

Ž .molecular weights, and one broad LMW peak 3383 .Note that this distribution also suggests the coexis-tence of, at least, two types of species in the solu-

Fig. 2. Molecular weight distributions obtained by GPC for the various asphaltene fractions derived from five different crude oils.

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000

53

tion. The LMW peak for the nC -asphaltenes is not9easily observed as in the previous two cases. Oneprominent peak with a peak molecular weight of7747 is observed with little tailing towards highmolecular weights. The shape of this peak suggeststhat in this case, the amount of LMW species ismuch smaller than in the case of nC -, and nC -5 7asphaltenes. This suggests rather strong interactionsamong asphaltene molecules such that very HMWspeciesraggregates are formed. The average molecu-lar weights obtained for the nC -, nC -, and nC -5 7 9asphaltenes were 2725, 6198, and 5607, respectively.

These studies also indicate that the precipitatedfractions contain a considerable amount of LMWspecies in solution. This is clearly indicated by thecontinuous distribution throughout the chromatogram

Ž .until the lower exclusion limit i.e. 100 for the setupused. When asphaltenes are dissolved in aromatic orpolar solvents, even at low concentration, there seems

Žto be a complex equilibrium: Rogacheva et al.,.1980; Pacheco-Sanchezand and Mansoori, 1998

MoleculeslMicelleslMicelle aggregates

Ž .In this equilibrium, the first species moleculesmay exhibit molecular weights up to 4000; the sec-

Ž .ond species micelles could have molecular weightsŽfrom 4000 to 10,000; and the third species micelle

.aggregates may possess molecular weights up to40,000,000. The existence of two distinct familiesdemonstrated in Fig. 2 are attributed to the asphal-tene molecule and asphaltene micelle.

5. SARA separation

The SARA separation was originally designed forŽ .characterization of residuals Jewell et al., 1974 .

The heavy end of a crude oil can be separated intofour distinct fractions namely saturates, aromatics,resins, and asphaltenes using the experimental SARAprocedure. A crude oil usually contains a consider-able amount of volatile material that must be re-moved prior to performing SARA separation of thecrude oil. This is accomplished by performing avacuum batch distillation at 10 mm Hg and roomtemperature until the weight of the residue reaches aconstant value. The first fraction to be separated

from the vacuum residue is the nC -asphaltenes. The5filtrate collected from the separation of the nC -5asphaltenes is commonly known as maltenes. It con-tains the remaining three fractions, saturates, aromat-ics and resins. These three fractions are separatedusing open-column liquid chromatography.

Saturate hydrocarbons on percolation in a n-pen-tane eluant, are not absorbed on activated silicaunder the conditions specified. The saturate fractionof the oil is eluted from the column with 1 l ofn-pentane at 5 mlrmin. The solvent is removedusing a rotary vacuum evaporator to recover thesaturates fraction.

Aromatic hydrocarbons are adsorbed on activatedsilica in the presence of n-pentane, and desorbed bytoluene after removal of the saturates under theconditions specified. The aromatic fraction of the oilis eluted from the chromatographic column usingtoluene at 5 mlrmin. The resin fraction of the oil iseluted from the chromatographic column using a90r10 toluenerMeOH solution at 5 mlrmin. Thisfraction is reported as wt.%. Solvent removal is to beperformed using standard laboratory procedures.

The bulk composition of crude oil A from the twoSARA separations performed on the vacuum residueis given in Table 3. Note from Table 3 that as muchas 21.05% by weight was separated from this crude

Ž .oil at 278C and 10 mm Hg vacuum distillation .These light ends were further analyzed by gas chro-

Ž .matography-mass spectroscopy GC-MS as will bediscussed later. Also note that the SARA separationperformed on the vacuum residue was successfullyaccomplished with an average recovery of about99%. The resins content of crude oil A is very small

Table 3Results from the SARA separation of crude oil A

Fraction wt.% of whole wt.% of vacuumcrude oil residue

Light compounds 21.0500"0.1150 100Žseparated at 278C

.at 10 mm HgSaturates 55.2369"1.6849 69.1949"1.3650Aromatics 19.9590"1.5373 25.2788"1.9488

Ž .Resins or polars 2.3242"0.0108 2.9437"0.0137Ž .Asphaltenes nC5 1.0833"0.0050 1.3753"0.0032Ž .Total recovery average 99.6534 98.7927

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000

54

Ž .compared to its total asphaltene content see Fig. 1 .The overall composition of this crude oil appears tobe a hostile environment for the stability of itsasphaltene content.

6. Diamondoids separation and characterization

Ž .Diamondoid adamantane has the same structureas the diamond lattice, highly symmetrical andstrain-free. It is generally accompanied by smallamounts of alkylated adamantanes. The isolation of asingle compound, adamantane, from such a complexmixture as petroleum is a consequence of its highly

Žunusual physical and chemical properties Vazquez.Gurrola et al., 1998 .

ŽAdamantane with freezing point of 2698C in.sealed capillary crystallizes in a face-centered cubic

lattice that is quite unusual for an organic compound.The special properties of adamantane are reflected inits mass spectrum that is also quite unusual.

Deposition of diamondoids can be particularlyproblematic during production and transportation ofnatural gas, gas condensates and light crude oils.These compounds are common in petroleum fluids,in general. However, their presence in crude oils isusually ignored due to their low concentration. Nev-ertheless, even small concentration of these com-pounds could lead to problems of deposition andplugging of flow paths. The high amount of gas

Žassociated with crude oil A gas-to-oil ratios2203 3.m rm suggests that diamondoids, if present in

appropriate amounts, may cause problem during itsproduction.

GC-MS analyses of diamondoids in petroleumfluids can be performed on the saturates fraction ofSARA separation, rather than on the whole crude oil.This is because the aromatic interferences need to beremoved by isolating the saturates fraction on silicagel. The light fraction of a crude oil, separated at278C and 10 mm Hg, must be also analyzed byGC-MS to determine whether it contains diamon-doids. Adamantane is expected to elute from the GCcolumn between nC and nC ; diamantane is ex-10 11pected to elute between nC and nC ; and tria-15 16mantane is expected to elute between nC and19nC . The same alkyl-substituted adamantanes found20in the saturated fraction of crude oil A were also

Table 4Alkyl substituted adamantanes found in crude oil A

Compound Mq Base peak RetentionŽ .time min

Ž . Ž .1-Methyl 150 14% 135 M-CH3 55.59adamantane

Ž . Ž .2-Methyl 150 68% 135 M-CH3 60.88adamantane

Ž . Ž .1,4-Dimethyl 164 15% 149 M-CH3 61.88adamantane

Ž . Ž .1,2-Dimethyl 164 15% 149 M-CH3 64.51adamantane

Ž . Ž .1-Ethyl 164 9% 135 M-C2H5 66.89adamantane

detected in this low-boiling fraction at similar reten-Ž .tion times and in higher concentrations see Table 4 .

7. Conclusions

The experimental procedures presented in thisreport provide the means for understanding theamounts and behaviors of the species that precipitateas well as the interactions among them. A detailedprocedure for the separation and analysis of thesuspended solids in the crude oils is presented. Theresults of the procedure can also reveal whether suchtests as cloud point, onset of flocculation of asphal-tene and other onsetrthreshold of deposition testscan be proceeded for the oil sample under study.

Total asphaltene content of crude oils are mea-sured using various precipitating agents and its rela-tion with the nature of the precipitating agent isestablished. This data along with the onsetrthresholddata may be used to tune theoretical modelsŽ .Kawanaka et al., 1991; Mansoori, 1997 of heavyorganics deposition. GPC studies of asphaltenes sep-arated from a crude oil will produce the size distribu-tion of asphaltene deposited as a function of thesolvent used. Such data, along with GPC size distri-

Žbution data Leontaritis and Mansoori, 1989; Vazquez.and Mansoori, in press for resins can be quite useful

for tuning theoretical polydisperse models of heavyŽorganics deposition Kawanaka et al., 1991; Man-

.soori, 1997 . The SARA separation of the heavy andlight fractions of a crude oil give the composition ofthe crude oil from the point of view of the four basic

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000

55

families of compounds, namely saturates, aromatics,resins and asphaltenes. Interactions among these fourfamilies of compounds have a profound effect on thestability of a crude oil that must be understood andmodeled in the stability analysis of crude oils. Char-acterization and concentration measurement of dia-mondoids can be particularly useful for petroleumreservoir fluids with high gas-to-oil ratios. The ana-lytic techniques presented in this report can be usedto find the composition and nature of such com-pounds present in a crude oil.

References

Escobedo, J., Mansoori, G.A., 1995. Viscometric determination ofthe onset of asphaltene flocculation: a novel method. SPEProd. Facil., 115–118, May.

Escobedo, J., Mansoori, G.A., 1997. Viscometric principles ofonsets of colloidal asphaltene flocculation in paraffinic oilsand asphaltene micellization in aromatics. SPE Prod. Facil.,116–122, May.

Ž .IP 359r82 ASTM D4007-87 , 1985. Determination of water andsediment in crude oil by centrifuge method. Standards forPetroleum and its Products. Institute of Petroleum, London,UK, pp. 359.1–359.8.

ŽIP 143r90 or ASTM D3279-90, 1985. Asphaltene heptane insol-.ubles in petroleum products. Standards for Petroleum and its

Products. Institute of Petroleum, London, UK, pp. 143.1–143.7.

Jewell, D.M., Albaugh, E.W., Davis, B.E., Ruberto, R.G., 1974.Integration of chromatographic and spectroscopic techniquesfor the characterization of residual oils. Ind. Eng. Chem.

Ž .Fundam. Vol. 13 3 , 278–282.Kawanaka, S., Park, S.J., Mansoori, G.A., 1991. Organic deposi-

tion from reservoir fluids. SPE Reservoir Eng., 185–192,May.

Leontaritis, K.J., Mansoori, G.A., 1989. Fast crude-oil heavy-component characterization using combination of ASTM,HPLC, and GPC methods. Int. J. Pet. Sci. Technol. 2, 1–12.

Letoffe, J.M., Claudy, P., Kok, M.V., Garcin, M., Vollet, J.L.,1995. Crude oils: characterization of waxes precipitated on

Ž .cooling by DSC and thermomicroscopy. Fuel 74 6 , 810–817.Mansoori, G.A., 1997. Modeling of asphaltene and other heavy

organics depositions. J. Pet. Sci. Eng. 17, 101–111.Pacheco-Sanchezand, J.H., Mansoori, G.A., 1998. Prediction of

the phase behavior of asphaltene micelleraromatic hydrocar-Ž .bon systems. Pet. Sci. Technol. 16 3–4 , 377–394.

Rogacheva, O.V., Rimaev, R.N., Gubaidullin, V.Z., Khakimov,D.K., 1980. Colloid J. USSR, 490–499.

Vazquez, D., Mansoori, G.A., in press. Procedures for Oil FieldSolid Deposits Analysis and Characterization.

Vazquez Gurrola, D., Excobedo, J., Mansoori, G.A., 1998. Char-acterization of crude oils from Southern Mexican oil fields.Proceedings of the EXITEP 98, International Petroleum Tech-nology Exhibition, Placio de Los Deportes, Mexico City,Mexico, D.F., 15th–18th November. .

Vuong, L.T., 1985. A simple and accurate experimental techniquefor measurement of the onset of asphaltene deposition frompetroleum reservoir fluids. MSc Thesis, Chemical EngineeringDepartment, University of Illinois at Chicago.

D. Vazquez and G.A. Mansoori Identification and Measurement of Petroleum Precipitates

J. Petrol. Sci. & Eng'g. 26(1-4): 49-55,2000