Beruflich Dokumente
Kultur Dokumente
luminate the strengths and weaknesses of each and to develop gel in hot toluene. Volatile components lost during the proc-
a standard method. An improved HPLC technique is de- ess were calculated by weight difference.
scribed and the results are applied to the problem of predicting
asphaltene stability, which has implications for many oil re- TLC-FID. An MK-5 Iatroscan (Iatron Labs Inc., Tokyo),
covery processes both through the possibility of deposition equipped with a flame ionization detector (FID), interfaced
and the potential for alteration of wettability. with a Model 203 PeakSimple data system (SRI Instrument),
was used to scan silica-coated quartz rods (Chromarod-SIII,
Experimental Materials and Methods Iatron Laboratories). The rods were 15.2-cm long and 1.0-mm
in diameter with a uniform coating of 5.0-m silica particles
Crude Oils. Six medium-gravity, dead crude oils, varying in (pore diameter 60 ). The FID detector was operated with a
API gravity from 22.6 to 37.2, were used in this study. Se- pure grade hydrogen flow rate of 160 ml/min, an air flow rate
lected physical and chemical properties are shown in Table 1. of 2.0 L/min supplied by a pump and a scan speed of 60s/scan.
Densities were measured using a Mettler/Paar DMA 40 with a Crude oil samples were dissolved in HPLC grade di-
circulating water bath for temperature control. API gravities chloromethane (DCM) at a concentration of 20 mg/mL. A 1
were calculated from measured densities. Average molecular L repeating syringe (Alltech Associate, Inc.) was used to
weight (MW) was determined by freezing point depression spot 10 to 20 g of sample on freshly activated rods. Devel-
(Precision Systems Cryoscope 5009). Kinematic viscosities opment steps included exposure to HPLC grade hexane for 30
were measured in Cannon-Fenske viscometers and converted min, HPLC grade toluene for 10 min, and a 95:5 mixture of
to absolute viscosity units. The amounts of asphaltene DCM and methanol for 4 min. The rods were dried in air for 3
precipitated by n-heptane (1 g oil:40 ml heptane) ranged from min between solvent exposures. A typical chromatogram is
1.3 to 8.7%. Refractive index (RI) was measured with an shown in Fig. 1 for S-Ven-39. Evaporative losses during the
Index Instruments GPR11-37 automatic refractometer. PRI is development steps were evaluated by tests of similarly treated
the RI of a mixture of oil and the least amount of heptane in preparative thin layer chromatography plates for which
which asphaltene aggregates can be observed microscopically changes in weight could be measured and by comparison of
(320X). The difference between RI of the oil sample and PRI peak sizes for undeveloped rods.
is a measure of asphaltene stability.13 Among these samples,
C-LH-99 had the most stable asphaltenes. Three of the oils (C-
R-00, S-Ven-39, and Tensleep-99) contained aphaltene aggre-
gates in the samples as received. In SQ-95 crystalline waxes
can be observed using crossed polar microscopy at ambient
conditions.
ring aromatic, DIPB, with a two-ring compound, 1- filtration through an 0.22 m filter, dried and weighed. Sev-
methylnaphthalene (1-MN) eluted as two well-resolved peaks. eral 0.5 ml aliquots of the maltene/hexane mixture were in-
Mixtures of the hexane solutions of DIPB, 1-MN, and C15 jected onto the HPLC columns, using a gas tight HPLC sy-
eluted as three distinct peaks (Fig. 2). ringe (HAMILTON).
methylnaphthalene Saturates and aromatics were eluted with hexane at a flow
rate of 1.5 ml/min. A typical chromatogram is shown in Fig.
3. The amounts of saturates and aromatics were calculated
diisopropylbenzene
from calibration factors measured with known compounds
(Table 2).
n-pentadecane
Table 2. RI Detector Calibration
Signal
Class Hydrocarbon
(area/mg)
saturates n-heptadecane 4,647
1-ring aromatics DIPB 7,088
2-ring aromatics 1-MN 12,517
Figure 2. Separation of paraffins from single-ring and double-ring 3-ring aromatics phenanthrene 16,022
aromatics.
Resins were strongly adsorbed and did not elute with hex-
ane. A 30% (v/v) dichloromethane/hexane backflush was
saturates used to elute the resins accumulated from at least three repli-
aromatics: cate injections to ensure that sufficient resins could be col-
single ring lected for accurate gravimetric determinations of the amount
of resin remaining after solvent evaporation.
The principal improvements in the HPLC method de-
scribed here over previously proposed HPLC techniques in-
clude:
the use of NH2-bonded columns,
2 rings analysis of the maltenes without solvent evaporation,
3 rings thus avoiding the uncertainties due to evaporation of
0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1
75 75
(a) A-95 (b) C-LH-99
50 50
25 25
0 0
Volatiles Saturates Aromatics Resins Asphaltenes Volatiles Saturates Aromatics Resins Asphaltenes
75 75
(c) C-R-00 (d) S-Ven-39
50 50
25 25
0 0
Volatiles Saturates Aromatics Resins Asphaltenes Volatiles Saturates Aromatics Resins Asphaltenes
75 75
(e) SQ-95 (f) Tensleep-99
50 50
25 25
0 0
Volatiles Saturates Aromatics Resins Asphaltenes Volatiles Saturates Aromatics Resins Asphaltenes
Figure 4. Comparison of SARA fractions determined by the ASTM method, by the improved HPLC method, and by TLC-FID.
SPE 75228 EVALUATING CRUDE OILS BY SARA ANALYSIS 5
The ASTM method also produces fractions of sufficient rial lost in the ASTM measurements was assumed to belong to
volume to permit measurements of fraction properties, such as the saturate fraction, the comparisons between saturates, aro-
RI, as shown in Fig. 5. RI of each oil is the volume-weighted matics, and resins determined by the ASTM and HPLC meth-
average of RI of each fraction. For the non-polar species that ods was quite close. (Note that the amounts of asphaltenes
dominate in a crude oil, RI is related to solubility parameters given for the ASTM and HPLC methods are from the same
and thus to solvent quality for asphaltenes. The low RI satu- measurements.) The major differences between the two tech-
rate fractions are poor asphaltene solvents whereas the high RI niques are the amounts of time, sample, and solvent required.
aromatics and resins are good solvents for asphaltenes. For both techniques there is a two day period required for
separation of the asphaltenes. Beyond that, however, the
1.65
Oil HPLC technique requires only about one hour per sample
Saturates compared to at least 20 hours for the ASTM separation. One g
Aromatics
1.60 of oil is adequate to give the same accuracy by the HPLC
Resins
technique as the 10 g required by the ASTM test. The solvent
1.55 requirements are on the order of 300 ml and 2 L per sample
for the HPLC and ASTM tests, respectively.
RI
1.50
SARA Fractions and Asphaltene Stability. The proportions
1.45 of each of the SARA fractions in a crude oil are related to the
stability of asphaltenes in that oil. Carbognani and Espidel14
1.40
demonstrated that reservoirs with asphaltene problems were
not primarily those with large amounts of asphaltene in the oil,
but those with high saturate fractions. Leontaritis15 recom-
1.35
A-95 C-LH-99 C-R-00 S-Ven-39 SQ-95 Tensleep-99 mended using the ratio of resins to asphaltenes as an indicator
of asphaltene stability, based on the hypothesis that resins con-
Figure 5. RI of oils and their saturate, aromatic, and resin frac- fer asphaltene stability by peptizing (or coating) asphaltenes.
tions.
Evidence is lacking to support either that hypothesis or the
utility of any specific ratio of resins to asphaltenes as an indi-
TLC-FID vs. ASTM. Summation of the volatile and saturate
cator of asphaltene stability. Similarly, Por16 (cited by Aso-
amounts from the TLC-FID method would significantly over-
maning and Watkinson17) defined a colloidal instability index
estimate the saturate fraction and underestimate the amount of
as the ratio of (asphaltenes+ saturates) to (resins+asphaltenes).
aromatics. Ignoring the volatilization and simply reporting the
As with the resin to asphaltene ratio, this index certainly
relative amounts of the four peaks would lead to even greater
should show some correlation with asphaltene stability, but no
divergence from the ASTM results. Only by separating all of
critical or definitive cut-off value can be expected.
the volatile material up to some fairly high boiling point
Both of these indices rely on an arbitrary differentiation
(Karlsen and Larter10 measured losses of approximately 50%
between asphaltenes and resins that is an analytical conven-
of n-C14 which boils at 253.5C, whereas the loss of n-C16
ience that does not correspond to a clear chemical distinction.
which boils at 286.8C was negligible) and analyzing that Recent studies have shown (1) that asphaltene stability can be
fraction separately, then combining the two sets of data would quantified by and predicted from measurements of the RI at
one be able to apply the TLC-FID technique to medium grav- the onset of asphaltene precipitation, PRI,18 (2) that broader
ity oil samples with any confidence, but the advantage of rapid distributions of asphaltenes are more stable than narrower
analysisthe main reason for using this techniquewould be ones, and (3) that resins confer stability by broadening the
lost. Apportioning the high molecular weight polar materials range of asphaltene properties.19 PRI for precipitation caused
between resins and asphaltenes, which is always based, rather by addition of n-C7 ranges from 1.4223 to 1.4513 RI units for
arbitrarily, on solvency, inevitably will vary from the ASTM the oils in this study. For a larger sample of oils, the average
and HPLC methods as well since the solvents used are not n-C7 PRI was 1.44 with a standard deviation of 0.02. One way
identical. On the basis of these comparisons, the TLC-FID to use the SARA fraction data to assess asphaltene stability
technique cannot be recommended for routine SARA analysis would be to estimate a value of RI from the volume fractions
of medium gravity oils. and compare that number to the expected onset for heptane (or
other alkanes, which are related to the heptane onset value18).
HPLC vs. ASTM. Differences between the improved HPLC The accuracy of the calculated value of RI can be evalu-
SARA fractions and those measured by the ASTM- ated by examining the correlation between SARA fraction
recommended method were mainly in the volatile and saturate data (acquired using the improved HPLC technique) and
fractions. There were no losses to evaporation in the HPLC measured values of oil RI, as shown in Fig. 6 for a set of RI
technique, unlike the ASTM method in which there were some and SARA data for 67 crude oils. The aromatic fraction has
losses (during evaporation of the solvent after removal of the little or no influence on RIoil whereas the saturates correlate
asphaltenes and in the various fraction recovery and solvent negatively (Fig. 6a) and the resins and asphaltenes (Fig. 6b)
evaporation processes). However, if all of the volatile mate- are positively correlated with RIoil.
6 T.G. FAN, J.X. WANG, J.S. BUCKLEY SPE 75228
1.48
RI oil (S * 1.4452 + A * 1.4982 + (R + As) * 1.6624 )/100 (1)
1.46
Saturates
y = 9E-05x + 1.4923
where S is the wt% of saturates, A is the wt% of aromatics, R
1.44 2
R =0 Aromatics is the wt% resins, and As is the wt% asphaltenes. In Fig. 7
measured RI values are compared with those calculated from
1.42
Eq. 1.
0 10 20 30 40 50 60 70 80
%
1.56
(a) saturates and aromatics
1.54 67 oils
1.56
y = 95x - 139
2
R = 0.4 1.52
1.54
Calculated RI
1.50
1.52
1.50 1.48
RI
y = 221x - 316
1.48 2 1.46
R = 0.6
1.46 1.44
Asphaltenes
1.44 1.42
Resins
1.42 1.44 1.46 1.48 1.50 1.52 1.54 1.56
1.42 Measured RI
0 5 10 15 20 25 30 35 40
% Figure 7. Comparison of measured and calculated values of RIoil
with estimates based on SARA analyses.
(b) resins and asphaltenes
Predicted and measured asphaltene stability can be com-
1.56
pared by calculating a value of RI = RIoil - PRI using an aver-
1.54 age PRI = 1.44. Measured and calculated values of RI =
RIoil - PRI are shown in Fig. 8 for the original set of six oils.
1.52 Exact boundaries for the stable and unstable asphaltene re-
gions cannot be defined, but oils with RI < 0.04 or 0.05 may
1.50
be problem oils whereas those with RI > 0.06 are more likely
RI
1.48
y = 317x - 455 to have stable asphaltenes.18 These boundaries are indicated
2
R = 0.7 by the shaded areas in Fig. 8. For two of the oils the measured
1.46 RI indicates that the asphaltenes are likely to be stable, one
Resins+Asphaltenes appears in the unstable region and the remaining three are in
1.44
the borderline region. Assessments of five of the six oils are
1.42
the same for the estimates based on the SARA fraction data.
0 10 20 30 40 50 In only one case, indicated by the circled data point, does the
% oil appear to be more stable than it really is.
0.07
and Bitumens Using Chemically Bonded Aminosilane, Anal.
Chem. (1986), Vol.58, No.12, 2389-2396.
0.06
6. Flix, G., Thoumazeau, E.,Colin, J.M., and Vion, G.: Hydrocar-
bon Groups Type Analysis of Petroleum Products by HPLC on
0.05
Specific Stationary Phases, J. Liq. Chrom. (1987), 10(10), 2115-
2132.
0.04
unstable 7. Chaffin, J.M., Lin, M.S., Liu, M., Davison, R.R., Glover, C.J., and
asphaltenes
Bullin, J.A.: The Use of HPLC to Determine the Saturate Con-
0.03
0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
tent of Heavy Petroleum Products, J.Liq. Chrom. & Rel. Tech-
nol. (1996), 19(10), 1669-1682.
RImeasured 8. McLean, J.D. and Kilpatrick, P.K.: Comparison of Precipitation
and Extrography in the Fractionation of Crude Oil Residua, En-
Figure 8. Asphaltene stability estimates compared to measured ergy & Fuels (1997) 11, 570-585.
values. RI = RIoil - PRI where RIoil estimates are made on the basis
9. Suzuki, Y: Studies on Quantitative TLC by means of a Hydro-
of SARA fraction data and Eq. 1 and PRI is taken to be 1.44 for the
purpose of estimation. gen FID. II Rapid Analysis of Fuel Oil Constituents, 21st An-
nual Meeting of the Japan Society for Analytical Chemistry 47
(1972). (in Japanese)
Conclusions 10. Karlsen, D.A. and Larter, S.R.: Analysis of Petroleum Fraction
Measurements of SARA fractions are highly dependent by TLC-FID: Application to Petroleum Reservoir Descrip-
on methodology. Results of different methods may not be tion,Org. Geochem. (1991) Vol.17, No.5, pp. 603-617.
comparable. Caution is needed in comparing results even 11. Vela, J., Cebolla, V.L., Membrado, L., and Andres, J.M.: Quanti-
from similar methods. tative Hydrocarbons Group Type Analysis of Petroleum Hydro-
The TLC-FID technique cannot be used for medium grav- conversion Products Using an Improved TLC-FID System, J.
ity oils without considerable additional analysis to ac- Chromatogr. Sci. (1995) Vol.33, 417-424.
12. Barman, B.N.: Hydrocarbon-Type Analysis of Base Oils and
count for components that boil at temperatures up to
other Heavy Distillates by Thin-Layer Chromatography with
250C. The apportioning of high molecular weight, polar, Flame-Ionization Detection and by the Clay-Gel Method,
aromatic material into asphaltenes and resins will not cor- J.Chromatogr.Sci. (1996) Vol.34, pp.219-225.
respond to the ASTM method. 13. Buckley, J.S., Hirasaki, G.J., Liu, Y., Von Drasek, S., Wang, J.X.,
An efficient HPLC methodology was developed and and Gill, B.S.: "Asphaltene Precipitation and Solvent Properties
shown to correspond closely to the more time-consuming of Crude Oils," Petroleum Sci. & Tech. (1998) 16, No. 3&4, 251-
ASTM-recommended method of SARA analysis. 285.
14. Carbognani, L. and Espidel, J.: Characterization of Solid Depos-
SARA fraction data can be used to screen crude oils for
its from Production Facilities. Identification of Possible Causes
likely asphaltene stability by relating a calculated RI of Deposits Formation, Vision Technologica (1995) 3, No. 1, 35-
based on the amounts of saturates, aromataics, and asphal- 42.
tenes plus resins to the RI conditions at the onset of as- 15. Leontaritis, K.J. and Mansoori, G.A.: "Asphaltene Flocculation
phaltene instability. During Oil Production and Processing: A Thermodynamic - Col-
loidal Model," paper SPE 16258 presented at the 1987 SPE Inter-
Acknowledgements nat. Symp. on Oilfield Chem., San Antonio, Feb 4-6.
This work was supported by the National Energy Technology 16. Por, N.: Stability Properties of Petroleum Products, Israel Inst. of
Laboratory (NETL) of the US DOE through contract DE- Petroleum and Energy, Tel Aviv, 1992.
AC26-99BC15204 and by support from industrial sponsors 17. Asomaning, S. and Watkinson, A.P.: Deposit Formation by
Asphaltene-Rich Heavy Oil Mixtures on Heat Transfer Surfaces,
including BP, Chevron, Gaz de France, IFP, Norsk Hydro, and in T.R. Bott (ed.) Understanding Heat Exchanger Fouling and its
TotalFinaElf. Crude oils were provided by ARCO, Chevron, Mitigation, Begell House, New York, 1999, pp. 283-290.
Shell, and the University of Wyoming. 18. Buckley, J.S. and Wang, J.X.: Crude Oil and Asphaltene Char-
acterization for Prediction of Wetting Alteration, Proc. 2000 In-
References ternat. Symp. on Reservoir Wettability, Socorro, 27-28 Sept. (in
1. Jewell, D.M., Weber, J.H., Bunger, J.W., Plancher, H., Latham, press JPSE)
D.R.: Ion-Exchange, Coordination, and Adsorption Chroma- 19. Wang, J.X. and Buckley, J.S.: Asphaltene Stability in Other
tographic Separation of Heavy-End Petroleum Distillates, Anal. Crude Oil Fractions, paper 49a presented at the 2002 Spring
Chem. (1972) 44, 1391-1395. AIChE Symposium, New Orleans, March 10-14.
2. Suatoni, J.C. and Swab, R.E.: Rapid Hydrocarbon Group-Type
Analysis by High Performance Liquid Chromatography, J.
Chrom. Sci. (1975), Vol.13, 361-366.