3618 Energy & Fuels 2007, 21, 3618–3621

Issues with Comparing SARA Methodologies

Abdel M. Kharrat,*,† Jose Zacharia,† V. John Cherian,‡ and Allwell Anyatonwu§
Schlumberger Canada Ltd., 9450-17 AVenue, Edmonton Alberta T6N 1M9, Canada; Dowell Schlumberger
Western S.A., P.O. Box 16818, Jebel Ali, Dubai, United Arab Emirates; and 16115 Park Row Suite 150,
Houston, Texas 77084

ReceiVed July 10, 2007. ReVised Manuscript ReceiVed August 31, 2007

One of the most common compositional analyses for petroleum samples is known as the SARA (saturates,
aromatics, resins, and asphaltenes) fractionation test. SARA fractionation is also used as one of the screening
criteria for asphaltene stability of reservoir fluids due to pressure depletion or commingling of different fluids.
There are numerous variations of this type of analysis. First, the extraction of asphaltenes is not consistent
from method to method. Asphaltenes are extracted using either pentane, hexane, or heptane. There are no
specific reasons for selecting one over the other, and usually the users do not associate differences in results
with the nature of the solvent. In addition, the extraction temperature could have an impact on the amounts of
asphaltenes extracted. The fractionation of maltenes is also a challenge, usually ignored by end-users. Assuring
no overlap between fractions and obtaining a very good mass balance are among these challenges. They could
be impacted by the type of packing material amount of solvents used for the chromatographic separation.
These SARA methods, referred to as standard methods, usually generate different results leading to confusion
if the users are not that familiar with analytical details of each method. This paper discusses the role of the
major parameters involved in generating the four fractions and how these parameters affect results, thus impacting
decision for the end-users. It also shows that it is impossible to perform any prediction of results when changing
from one method to another.

Introduction
Characterization of fluids is required for many reasons as
given in the literature. The fluid behavior inside the reservoir
depends on composition. Modeling this behavior with time
requires compositional analysis along with other physical
parameters.1 Proper reservoir management necessitates the
knowledge of reservoir conditions, such as pressure and
temperature in addition to composition.2 During transportation
and storage, the mixing of different fluids could cause perturba-
tion of the fluids system. The presence of noncompatible fluids
leads to precipitation and deposition. The refining process is
also dependent on the nature of the fluids. One of the most
common methods is the separation of oil into four fractions:
saturates, aromatics, resins, and asphaltenes (known as SARA).
Jewel et al.3 were the first researchers who worked out this
type of fractionation method. Development of other approaches
* To whom correspondence should be addressed. E-mail:
akharrat@slb.com.
† Schlumberger Canada Ltd.
‡ Dowell Schlumberger Western S.A.
§ 16115 Park Row Suite 150.
(1) Pina, A.; Mougin, P.; Behar, E. Oil Gas Sci. Technol. 2006, 61,
319–343.
(2) Karan, K.; Hammami, A.; Flannery, M.; Stankiewicz, A. Systematic
Evaluation of Asphaltenes Instability and Control During Production of Live
Oils: A flow Assurance Study. Pet. Sci. Technol. 2003, 21, 629–645.
(3) Jewell, D. M.; Weber, J. H.; Bunger, J. W.; Plancher, H.; Latham,
D. R. Anal. Chem. 1972, 44, 1391–1395.
(4) Suatoni, J. C.; Swab, R. E. J. Chromatogr. Sci. 1975, 13, 361–366.
(5) Miller, R. Anal. Chem. 1982, 54, 1742–1746.
(6) Radke, M.; Willish, H.; Welte, D. H. Anal. Chem. 1984, 56, 2538–
2546.
(7) Grizzle, P. L.; Sablotny, D. M. Anal. Chem. 1986, 58, 2389–2396.
(8) Chaffin, J. M.; Lin, M. S.; Liu, M.; Davison, R. R.; Glover, C. J.;
Bullin, J. A. J. Liq. Chromatogr. Relat. Technol. 1996, 19, 1669–1682.
was the focus of later studies.4–9 This effort led to the
development of the ASTM D 2007 method,10 a rather incon-
venient method because it consumes large quantity of sol-
vents, adsorbent, and oil. Consequently, other methods were
developed.8,11–13 During the Fifth International Conference on
Petroleum Phase Behavior and Fouling at Banff in 2004, a
discussion was held on standardizing a method for asphaltene
determination. The discussion led to no agreement because of
the complexity of the available methodologies and lack of
understanding of correlation among them. As a consequence,
we undertook this study to investigate the possibility of
establishing a correlation between methods and studying the
impact of the major method parameters on the results. Three
laboratories were involved in this work to study the effect of
most of the parameters.
Experimental Section
Data from three laboratories (referred to as Lab 1, Lab 2, and
Lab 3) using three different methods for SARA fractionation are
presented here.
Sample Size. Lab 1 uses 2 g of oil for asphaltene precipitation
and 300 mg of maltenes for SAR separation. Lab 2 uses 2 g of oil
for asphaltene precipitation and 350 mg of maltenes for SAR
(9) Felix, G.; Thoumazeau, E.; Colin, J. M.; Vion, G. J. Liq. Chromatogr.
1987, 10, 2115–2132.
(10) ASTM D 2007. Standard Test Method for Characteristic Groups
in Rubber Extender and Processing Oils by Clay-Gel Adsorption Chroma-
tography Method, 1993.
(11) Fan, T.; Buckley, J. S. Energy Fuels 2002, 16, 1571–1575.
(12) Fuhr, B. J.; Hawrelechko, C.; Holloway, L. R.; Huang, H. Presented
at the 5th International Conference on Petroleum Phase Behavior and
Fouling, Banff, Canada, June 13–17, 2004.
(13) Carbognani, L.; Buenrostro-Gonzalez, E. Energy Fuels 2006, 20,
1137–1144.

10.1021/ef700393a CCC: $37.00  2007 American Chemical Society

Published on Web 10/23/2007
SARA Methodologies Energy & Fuels, Vol. 21, No. 6, 2007 3619

Table 1. Summary of the Analytical Parameters Performed by the Three Laboratories

parameter Lab 1 Lab 2 Lab 3
sample size
for asphaltenes 2 g of oil 2 g of oil 1 g of oil
for SAR 300 mg of maltenes 350 mg of maltenes 200 mg of maltenes
topping concentrator spinning band no topping
aspahltene 1:40 heptane 1:30 heptane 1:30 pentane
precipitation 98 °C room temperature 39 °C
asphaltene filtration 0.45 µm filter 0.45 µm filter 2.5 µm filter
asphaltene washing hot heptane heptane at RTa pentane at RTa
asphaltene extraction dichloromethane chloroform chloroform
column chromatography alumina activated at 430 °C alumina and silica activated at 250 °C alumina and silica activated at 180 °C
chromatography
saturates 250 mL of heptane 140 mL of heptane 70 mL of pentane
aromatics 250 mL of toluene 180 mL of 2:1 heptane–toluene 70 mL of 1:1 pentane–dichloromethane
resins 250 mL of 1:1 dichloromethane– 60 mL of 1:1:1 toluene–chloroform– 70 mL of methanol
methanol methanol
solvent removal rotary evaporator hot plate hot plate
a RT: room temperature.
separation. Lab 3 uses 1 g of oil for asphaltene precipitation and Table 2. Solvent Effect on Asphaltene Precipitation
200 mg of maltenes for SAR separation.
Topping. Lab 1 uses solvent evaporator (nonrotating) at 80° and % asphaltene % asphaltene
at a vacuum of 26.5 in. of mercury while Lab 2 uses spinning band crude precipitated precipitated asphaltene
oil with pentane with heptane ratio C5/C7
distillation and Lab 3 does not perform any sort of topping.
Asphaltene Precipitation and Filtration. Lab 1 precipitates oil 1 20.2 10 2.0
asphaltene using heptane (1:40 ratio) by refluxing at the boiling oil 2 8.8 4.1 2.2
point of heptane for 2 h. Lab 2 also precipitates with heptane oil 3 25.0 17.1 1.5
oil 4 10.5 0.04 263
(1:30 ratio) but at room temperature, whereas Lab 3 precipitates
oil 5 3.7 0.04 94
asphaltenes with pentane (1:30 ratio) by refluxing at the boiling
point of pentane followed by an incubation time of 24 h.
For filtration, Labs 1 and 2 use 0.45 µm pore size filters whereas Table 3. Effect of Washing on Percentage Asphaltene
Lab 3 uses a 2.5 µm filter. Calculated
Asphaltene Washing. Lab 1 uses hot heptane washing using a % asphaltene calculated with
Soxhlet extractor. Lab 2 uses heptane at room temperature until crude oil room temperature washing heptane Soxhlet washing ratio
solvent is clear. Lab 3 uses normal pentane at room temperature.
Asphaltene Extraction. Lab 1 uses dichloromethane to extract oil 6 2.76 2.43 1.1
oil 7 8.80 8.67 1.0
the asphaltene from the filter paper using a Soxhlet extractor. Labs
oil 8 0.44 0.40 1.1
2 and 3 use chloroform to extract asphaltene deposited on the filter. oil 9 0.64 0.38 1.7
Chromatography Column. Lab 1 uses a stainless steel (SS) oil 10 1.96 0.37 5.3
column (1 cm × 60 cm) packed with alumina activated at 430 °C oil 11 2.05 1.5 1.4
for 24 h. Lab 2 uses an SS column (1 cm × 110 cm) packed with oil 12 18.2 12.1 1.5
alumina and silica gel activated at 250 °C for 24 h. Lab 3 uses a oil 13 1.60 0.30 5.3
glass column (1 cm × 50 cm) packed with alumina and silica gel
activated at 180 °C for 24 h. Table 4. SARA Fractions for Oils 14 and 15
Chromatographic Elution. Labs 1 and 2 use a pump to deliver
solvent at the rate of 2 mL/min. Lab 3 uses manual gravity drain SARA % % % % %
crude oil method saturates aromatics resins asphaltenes recovery
method.
For eluting saturates Lab 1 uses 250 mL of heptane whereas oil 14 Lab 1 25.6 23.6 38.9 10.8 98.9
Lab 2 uses 140 mL of heptane. Lab 3 uses 70 mL of pentane to Lab 2 23.7 33.9 25.5 16.6 99.6
elute saturates. Lab 3 25.7 39.1 11.6 12.1 88.5
oil 15 Lab 1 61.2 18.8 17.6 0.10 97.7
Lab 1 uses 250 mL of toluene to elute the aromatic fraction,
Lab 2 60.2 25.3 13.5 0.8 99.9
while laboratory 2 uses 180 mL of a mixture of heptane and toluene Lab 3 49.7 23.5 9.2 0.3 82.7
in the ratio 2:1. Lab 3 uses 70 mL of a mixture of pentane and normalized 60.1 28.4 11.1 0.4 100
dichloromethane in the ratio 1:1. Lab 3 C 61.1 26.3 9.2 0.3 96.9
For resins, Lab 1 uses 250 mL of a mixture of dichloromethane
and methanol in the ratio 1:1. Lab 2 uses 60 mL of a mixture of
toluene, chloroform, and methanol in the ratio 1:1:1, and Lab 3 and extract asphaltenes from oils and bitumen are pentane and
performs resin elution with 70 mL of methanol. heptane. It is well established in the literature14 that the higher
Lab 1 uses a final column flush with 100 mL of dichloromethane the carbon number of the solvent, the less asphaltenes are
followed by 100 mL of methanol. This fraction is added up to the precipitating. Because both solvents are being used by labora-
resin fraction. This step is not followed by Labs 2 and 3. tories, users of the data are always interested in how to interpret
All these parameters are tabulated in Table 1. the data from two different laboratories, one using pentane and
Recovery of Separated Fractions. After chromatography separa-
tion, each fraction needs to be dried to remove the solvent and weighed
the second using heptane.
to calculate the final percentage composition. Lab 1 used rotary We determined asphaltenes using heptane and pentane for
evaporator under vacuum followed by hot plate with nitrogen purging different oils having different asphaltene contents. We computed
to dry the fractions. Labs 2 and 3 used a hot plate for the drying. the ratios of amount of asphaltenes generated by pentane over
the amount of asphaltenes generated by heptane. The results
Results and Discussion are presented in Table 2.

Asphaltene Precipitation and Extraction. Correlation (14) Speight, J. G. The Chemistry and Technology of Petroleum; Marcel
between SolVents. The most common solvents used to precipitate Dekker: New York, 1999.
3620 Energy & Fuels, Vol. 21, No. 6, 2007 Kharrat et al.

Table 5. Effect of Solvent Amount on Separation and Recovery Using the Lab 1 Method
% saturates with % aromatics with % resins % recovery
250 mL 100 mL 250 mL 100 mL 250 mL 70 mL
crude oil (method 1) (method 2) (method 1) (method 2) (method 1) (method 2) method 1 method 2
oil 16 72.7 71.8 15.8 15.4 10.6 4.7 99.1 92.0
oil 17 85.1 84.4 8.2 8.2 5.8 0.2 99.0 92.8
oil 18 49.7 48.4 18.0 17.0 27.3 14.1 97.4 81.9
oil 19 59.7 45.7 26.5 27.7 14.0 12.2 98.8 85.6

It is expected to have asphaltenes content higher using pentane
than heptane,14 but the ratios shown here are not constant. It
was reported that these ratios are in the range of 1.2–1.2515
and 1.38,16 much lower in some cases than ours. These ratios
are oil dependent, and it is not possible to extrapolate them from
the literature. The lack of correlation is the first proof that
standardization based on analytical methodology will be very
hard to accomplish. When comparing asphaltene content data
generated from different methodologies, the users should pay
attention to this parameter and not use the ratio published in
the literature.
Effect of Washing. Washing is rinsing the asphaltene fraction
with the precipitating solvent to eliminate the oil that could be
trapped inside the asphaltene cake during filtration. Since there
are numerous precipitation methods, it is expected that washing
will have different effects. Many methods describe this step as
rinsing the asphaltenes filtered until the solvent going through
the filter becomes clear. Calculating the asphaltene content from
the weight of unwashed filter cake is a common practice in the
industry. Even though this saves a lot of time, the accuracy of
the method is jeopardized. An independent study was done on
10 crude oils with varying asphaltene content to understand the
criticality of the washing. The C7 asphaltene content measured
with Soxhlet washing with hot heptane as outlined in the
Experimental Section (Lab 1 method) is compared with the one
without washing in Table 3.
Anderson and Stenby17 when studying the temperature effect
on precipitation of asphaltenes found a ratio of 1.33 between
24 and 80 °C, after performing washing. In our case, ratios were
between 1 and 5.3, presented in Table 3, leading to believe that
extraction of trapped maltenes in the asphaltene cake is more
efficient at the reflux temperature of heptane than at room
temperature.
Saturates Aromatics and Resins Extraction. Saturates,
aromatics, and resins (SAR) results for two crude oil samples
performed with three different approaches as outlined in the
Experimental Section are presented in Table 4 for comparison.
Oil 14 has low volatiles (0.6% topping) compared with oil 15
(14.2% topping).
Lab 3 normalized results are calculated to force recovery to
100%. This is common practice, and it is easy to detect because
the recovery would be always 100.00%, impossible to achieve
in all the cases.
Lab 3 C results were corrected on the basis of the fact that
they lost 14.2% volatiles, and those volatiles were assumed to
be 80% saturates and 20% aromatics.
Effect of Topping. Topping is the removal of volatiles by
means of distillation, rotary evaporation, or heating under
atmospheric or reduced pressure. Topping should be performed
at a temperature and pressure that do not allow further
(15) Buenrostro-Gonzalez, E.; Lira-Galeana, C.; Gil-Villegas, A.; Wu,
J. AIChE J. 2004, 50, 2552–2570.
(16) Alboudwarej, H.; Akbarzadeh, K.; Beck, J.; Svrcek, W. Y.;
Yarranton, H. W. AIChE J. 2003, 49, 2948–2956.
(17) Anderson, S. I.; Stenby, E. H. Fuel Sci. Technol. Int. 1996, 14,
261–287.
Table 6. Normalization of Results for Oils 20 and 21
crude oil % saturates % aromatics % resins % recovery
oil 18 48.4 17.0 14.1 81.9
oil 18 normalized 59.1 20.7 17.2 100
oil 19 45.7 27.7 12.2 85.6
oil 19 normalized 53.4 32.4 14.2 100
hydrocarbons to be lost in the saturates, aromatics, and resins
analysis when the solvent is removed.
The method used in Lab 3 produced low recovery for both
oils. This is because under the Lab 3 method topping is not
performed, and hence some hydrocarbons are lost during the
evaporation of the solvents after fractionation. This is evident
from the low saturate value for oil 15 using the Lab 3 method.
Since Labs 1 and 2 performed topping, both laboratories
produced comparable results for saturates and obtained a higher
recovery. These results show clearly that the separation of
volatiles is a critical step to perform during SARA analysis. It
is important though to highlight that spinning band distillation
would be preferred to the rotary evaporator or concentrator as
these techniques do not allow collection of the light fraction,
and consequently it would be impossible to characterize it.
Effect of SolVent and Packing Material. The nature and
amount of solvent and the packing material have an impact on
the column separation.
Since saturates are nonpolar, their interaction with the
adsorbent is minimal. Differences in saturates recovery is
attributed to the difference in solvent (type and quantity). Labs
1 and 2 used heptane in 250 and 140 mL quantities while Lab
3 used only 70 mL of pentane. For these two oils, it seems that
140 mL of heptane was sufficient enough to extract all saturates
from both oils; when 70 mL of pentane is used, all saturates
seemed to be recovered after correction due to topping.
The recoveries for the aromatic fractions were not consistent.
Labs 2 and 3 had higher recovery than Lab 1 probably due to
the lack of separation between the aromatics and the resins.
Lab 3 obtained low recovery of resins. Lab 3 used 70 mL of
methanol to elute this fraction. This amount is most probably
not enough to elute the entire fraction of resins. Solvent amount
and packing material both could be the origin of these
differences. We decided to study the impact of the amount of
eluting solvent on four oils using the same packing material.
The results are presented in Table 5.
For oil 19, 100 mL of heptane was not enough to extract the
entire saturate fraction. For oil 18, 70 mL was not enough to
extract the entire resin fraction. These results impacted the
recovery. Only 81.9% and 85.6% recoveries were obtained for
these two oils.
The practice of normalization, a common practice, generates
erroneous results.18 Method 2 in Table 5 has 81.9% and 85.6%
recovery for oils 18 and 19, respectively. The normalized results
are presented in Table 6. Forcing the recovery to 100%
(18) Kharrat, A. M. Presented at the 5th International Conference on
Petroleum Phase Behavior and Fouling, Banff, Canada, June 13–17, 2004.
SARA Methodologies
drastically changes the results. The change is also oil- and
recovery-dependent.
Conclusions
SARA methodologies do not generate similar results. Com-
paring results from different methodologies is very risky and
could lead to erroneous conclusions.
There are a few points to watch for when reading SARA
results:
Energy & Fuels, Vol. 21, No. 6, 2007 3621
1. The recovery should be close to 100%.
2. Sample topping should be performed initially to avoid loss
of volatiles during solvent removal. This helps to achieve a mass
balance close to 100%. It is recommended to use a spinning
band distillation to perform this step.
3. Solvents should be used in quantities that do not allow
low recoveries.
4. Correlation between methods is shown to be very difficult.
EF700393A