Full text of DOI:10.1007/s11746-001-0307-y

Screening Vegetable Oil Alcohol Esters
as Fuel Lubricity Enhancers
a,
a
b
D.C. Drown *, K. Harper , and E. Frame
a
b
Department of Chemical Engineering, University of Idaho, Moscow, Idaho 83844-1021, and U.S. Army
Tank-Automotive Research, Development, and Engineering Center Fuels and Lubricants Research Facility
Southwest Research Institute), San Antonio, Texas 78228-0510
(
ABSTRACT: Methyl and ethyl monoalkyl esters of various veg- was developed to predict lubricity of fuels intended for ground
etable oils were produced for determining the effects of type of equipment (4). Generally, a maximum wear scar of 450 µm in
alcohol and fatty acid profile of the vegetable oil on the lubricity
of the ester. Four methyl esters and six ethyl esters were analyzed
for wear properties using the American Society for Testing and
Materials method D 6079, Evaluating Lubricity of Diesel Fuels
by the High-Frequency Reciprocating Rig. Ethyl esters showed
noticeable improvement compared to methyl esters in the wear
properties of each ester tested. No correlation was found be-
tween lubricity improvement and fatty acid profile of the ester,
except that esters of castor oil had improved lubricity over other
oils with similar carbon chain-length (C ) fatty acids.
the HFRR test indicates acceptable fuel lubricity (5). The ob-
jective of this project was to define the lubricity improvement
of JP-8 fuel that was imparted by biodiesel ester fuel samples
prepared with varying chemical compositions.
MATERIALS AND METHODS
Transesterification reaction. The monoalkyl esters were made
by a transesterification reaction carried out at room tempera-
ture and atmospheric pressure in approximately 2500-mL
batches. Sodium methoxide or sodium ethoxide (depending on
the alcohol being used) was used as a catalyst at a concentra-
tion of 0.75% sodium by weight of triglyceride. To shift the re-
action toward monoalkyl esters, 100% excess alcohol (6 mol)
was used to obtain high (88–92%) conversions of oil into ester.
18
Paper no. J9696 in JAOCS 78, 579–584 (June 2001).
KEY WORDS: Biodiesel fuel, engine wear, fatty acid, lubric-
ity, monoalkyl esters, vegetable oil.
Biodiesel ester fuels have been examined as an alternative to The unpurified ester was then decanted off the top and subse-
petroleum fuels and as an additive to improve the lubricity of quently purified by a patented two-stage extraction process
these fuels. Previously tested biodiesel ester fuel samples with fresh glycerine (certified A.C.S. 99.7%; Fisher Scientific,
have given improved lubricity test results, but attention was Fairlawn, NJ) (6). The glycerine extraction/purification process
not focused on the vegetable oil from which the fuel was de- was chosen in order to reduce the moisture content of the final
rived (1). Each vegetable oil (and even specific varieties) has product. Traditional water washing leaves a high Karl Fischer
a characteristic fatty acid content profile. High-erucic acid moisture content in the ester, which creates corrosion problems
rapeseed (HEAR) is raised as an industrial oil crop and is not and hence is unacceptable for military applications.
an edible vegetable oil. One of the industrial uses of HEAR
Materials. The vegetable oils used in the transesterifica-
and its esters is as a component in specialty lubricants. Test- tion reaction were soy (generic brand salad oil labeled 100%
ing the ester of an oil with a known fatty acid profile would soy oil, bulk restaurant carboy from local wholesale grocery
help determine if the fatty acid profile influences lubricity store, Clarkston, WA), Sterling rapeseed (local farm 1998
characteristics of its ester. Additionally, using ethanol or crop grown under contract to Idaho TransTech, Inc., Moscow,
methanol as the alcohol in the esterification reaction would ID, custom crushed by Montana Specialty Mills, Great Falls,
help in analyzing the effects of the type of alcohol on improv- MT), Dwarf Essex rapeseed (local farm 1997 crop, custom
ing lubricity of the ester.
crushed by University of Idaho Agricultural Engineering
The U.S. Army has adopted a Single Fuel Forward policy, Dept., Moscow, ID), coconut (generic brand popcorn oil la-
in which aviation-grade turbine-engine fuel (JP-8) (2) is used beled 100% coconut oil, bulk restaurant carboy from local
in Army ground vehicles and equipment. The fuel lubricity of wholesale grocery store, Clarkston, WA), castor (bulk tank
many JP-8 fuels and some low-sulfur diesel fuels is not ade- truck sample of #1 castor oil donated by Lifelast, Inc., Van-
quate to provide wear protection for some ground equipment couver, WA), and partially hydrogenated canola (demonstra-
fuel injection pumps (3). The JP-8 specification contains the tion sample from French fry potato processor discontinued
Standard Ball-on-Cylinder Lubricity Evaluator (BOCLE) pilot test run donated by Idaho TransTech, Inc., Moscow, ID).
ASTM D5001 lubricity test. Passing this test does not guaran- The two different types of alcohol used were ethanol (USP
tee acceptable fuel lubricity for ground application fuel system absolute-200 proof, AAPER Alcohol & Chemical Co., Shel-
components. The high-frequency reciprocating rig (HFRR) test byville, KY) and methanol (certified A.C.S. 99.9%, Fisher
Scientific). All esters were made using elemental Na (lump
*
To whom correspondence should be addressed.
9
9%, aldrich Chemical Co., Milwaukee, WI) reacted with al-
E-mail: ddrown@uidaho.edu
Copyright © 2001 by AOCS Press
579
JAOCS, Vol. 78, no. 6 (2001)

580
D.C. DROWN ET AL.
TABLE 1
JP-8 Base Fuel Properties
to JP-8 fuel. Fuel blends containing biodiesel were evaluated
by the HFRR lubricity test to determine if the biodiesel ester
fuel additive gave improved lubricity to the fuel.
The HFRR (PCS Instruments, London, England) uses an
electromagnetic vibrator that oscillates a moving specimen
against a fixed specimen immersed in the test fuel over a
small amplitude (7). The HFRR standard operating conditions
are as follows (8,9): fluid volume (mL), 2; applied load (kg),
a
JP-8
AL-24666
MIL-DTL-83133E
requirementsd
Methodb
c
Property
API gravity (60°F)
Density (15°C, kg/L)
Flash point (°C)
Freeze point (°C)
Color
D 1298
D 12980
D 93
D 2386
D 1500
D 5291
47.9
37–51
0.7884 0.775–0.840
48
−61.5
<0.5
14.23
38 minimum
−47 maximum
Report
0
.2; speed (Hz), 50; duration (min), 75; fluid temperature
Hydrogen (mass %)
Net heat of combustion
13.4 minimum
2
(°C), 25 or 60; stroke (mm), 1.0; bath surface area (cm ), 6;
repeatability (mm), 0.062; reproducibility (mm), 0.127. The
HFRR data reported are the means of two or three test runs.
The test repeatability standard is 0.080 mm, and the operator
runs at least two tests to determine if they repeat within 0.080
mm; if not, then a third test is run.
(
MJ/kg)
Total acid number
mg/g KOH)
D 240
43.2
42.8 minimum
(
D 664
D 4294
D 6079
0.01
0.07
0.770
0.015 maximum
0.30 maximum
Not required
Sulfur (mass %)
HFRR (mm)
Scuffing load wear test (g) D 6078
Distillation (°C for
percentage off)
Initial
2100
Not required
Kinematic viscosity measurements were made according
to ASTM D 445 (8). Gas chromatography was used to deter-
mine the ester fatty acid profile, free glycerine content, and
free alcohol content. The fatty acid profile was determined by
using an HP5890 Series II gas chromatograph with an HP
D 86
165
173
175
181
192
206
Report
205 maximum
Report
Report
Report
10%
20%
50%
90%
7673 autosampler (Hewlett-Packard Co., San Fernando, CA),
End point
300 maximum
flame-ionization detector, and a 0.25 mm × 30 m, 0.25 µm
film thickness DB-23 capillary column (J&W Scientific, Fol-
som, CA). The free glycerine content and free alcohol con-
tent were also determined using the HP5890 gas chromato-
graph and autosampler, but with a 0.25 mm × 30 m DB-1 cap-
illary column (J&W Scientific). The free glycerine method
was adapted from one published by the Research Institute for
aJP-8 fuel, aviation-grade turbine engine fuel; API, the density-reporting units
specified by the military fuel specification; HFRR, high-frequency recipro-
cating rig.
b
All methods from American Society for Testing and Materials (8).
c
Sample provided by Southwest Research Institute, San ANtonio, TX.
d
Reference 2; Report = report to customer, no required range.
cohol to form sodium methoxide or sodium ethoxide as the Chemistry and Technology of Petroleum Products (10). The
catalyst in the respective alcohol solution. Southwest Re- free alcohol content was determined by spiking a sample with
search Institute (SwRI; San Antionio, TX) sample AL-24666 a known amount of alcohol to calibrate the analysis method.
of military grade aviation fuel JP-8, whose properties are
shown in Table 1, was used as the blending agent.
RESULTS AND DISCUSSION
Analytical methods. Lubricity testing was done according
to the HFRR (60°C) method D 6079 (8) with additions of Fatty acid profile. The results of the fatty acid profile analy-
monoalkyl esters of biodiesel at 0.1, 0.5, and 1.0% by volume ses are given in Table 2. The variety in the number of carbons
TABLE 2
Fatty Acid Profiles of Ester Given in Weight Percentage of Ester
Fatty acida
Ester
Caprylic
8:0)
Capric
(10:0)
Lauric
(12:0)
Myristic
(14:0)
Palmitic
(16:0)
Stearic
(18:0)
Oleic
(18:1)
Linoleic
(18:2)
Linolenic
(18:3)
(
Coconut ethyl
Soy methyl
Soy ethyl
7.23
—
—
5.72
—
—
46.03
—
—
18.57
—
—
9.54
10.66
10.50
2.99
4.44
4.36
7.38
23.57
23.15
2.32
52.25
52.16
—
7.03
7.12
Palmitic
Stearic
(18:0)
Oleic
(18:1)
Linoleic
(18:2)
Linolenic
(18:3)
Ricinoleic
(18:0)(–OH)
Arachidic
(20:0)
Arachidonic
(20:1)
Erucic
(22:1)
(16:0)
b
Canola ethyl
4.93
3.22
3.27
2.70
2.72
1.15
1.14
3.90
1.33
1.33
0.98
1.02
1.17
1.17
82.01
17.75
17.87
13.69
13.75
4.24
4.72
12.79
12.96
10.77
10.84
5.26
0.58
5.74
5.90
7.08
7.12
0.55
0.55
—
—-
—
—
—
—
0.95
9.91
9.86
7.37
7.34
0.40
0.43
0.20
42.74
42.41
50.95
50.63
—
c
Sterling methyl
0.96
0.96
0.81
0.84
—
c
Sterling ethyl
c
Dwarf Essex methyl
Dwarf Essex ethyl
c
Castor methyl
Castor ethyl
a
87.19
87.19
3.68
5.26
—
—
Fatty acid (# carbon: # double bonds).
b
From partially hydrogenated canola oil.
From high-erucic acid rapeseed varieties.
c
JAOCS, Vol. 78, no. 6 (2001)

ESTERS AS FUEL LUBRICITY ENHANCERS
581
TABLE 3
Other Physical Characteristics
in the castor esters were roughly 23 times higher, with a mean
of 1.16%. The castor esters also had a much higher viscosity.
Kinematic viscosity. As with the free glycerine analysis,
all samples were within a kinematic viscosity range of 3–6
cSt except for the castor samples (Table 3). They were higher,
but this time only by a factor of about 3. This difference in
kinematic viscosity and free glycerine content led to treating
the esters of castor oil as a separate subset as described in the
section on lubricity testing.
Free alcohol content. The free alcohol contents of the
methyl esters were all undetectable (below 0.03%) except for
the castor methyl ester at 0.95% (Table 3). The ethyl esters
ranged from 0.07% in the Dwarf Essex ethyl ester to 2.32%
in the castor ethyl ester. All of these values except castor were
also near the lower detectable limits of the testing procedure
used, as in the free glycerine analysis.
Free
glycerine
(wt%)
Kinematicb
viscosity
(cSt)
Alcohol
content
(wt%)
a
Ester
Coconut ethyl
Soy methyl
Soy ethyl
Canola ethyl
Sterling methyl
Sterling ethyl
Dwarf Essex methyl
Dwarf Essex ethyl
Castor methyl
Castor ethyl
0.07
0.04
0.04
0.07
0.03
0.06
0.04
0.03
1.21
1.10
2.91
4.55
4.46
5.46
5.72
6.01
6.22
6.48
12.69
17.63
0.12
ND
ND
0.15
ND
0.25
ND
0.07
0.95
2.32
a
Method modified from Reference 10.
Measurements made according to ASTM D 445.
b
Lubricity testing. The results of the lubricity tests were an-
in the carbon chain of each fatty acid is easy to see, as is the alyzed by dividing the test population into three different sub-
number of double and OH bonds. Another important feature sets. The first subset was the castor oil. Castor oil has a unique
to note is the similarity in the profiles between the methyl and OH group in the middle of the 18-carbon chain of ricinoleic
ethyl esters of the same vegetable oil. This showed that the acid, which amounts to approximately 90% of the fatty acid
alcohol did not affect the fatty acid profile of the ester, so the profile (Table 2). Owing to this difference in chemical struc-
influence of the profile and alcohol type could be separated ture and to the increased free glycerine contents of these sam-
and evaluated independently.
ples, it was decided to treat castor oil as its own subset. The
Free glycerine content. Free glycerine was low in most second subset was the methyl esters, and the third was the
samples (Table 3). For all samples except castor, the mean of ethyl esters.
0
.05% was very close to the 0.03% lower detectable limit of
In all cases, the lubricity of the sample tested improved
the analytical procedure. This may have caused some error with increasing volume of ester additive (Fig. 1), where the
because the amount of free glycerine in the sample was so declining wear from 0.1 to 0.5% ester additive was clearly
close to the end of the detectable range. Although 0.05% free seen. The same trend continued with the lowest wear at 1.0%.
glycerine is above the 0.02% ASTM PS 121-99 provisional In the case of 0.1% ester additive, there was no statistical dif-
biodiesel specification (8), when the esters are blended at 1% ference in the lubricity of any subset compared to variations
or less in JP-8 the resulting free glycerine level is significantly in the pure JP-8 turbine fuel. When all 0.1% ester samples
below the biodiesel specification, and the military specifica- were used in the determination of the statistical parameters,
tion has no glycerine limit. The free glycerine concentrations the results showed a normal distribution (Fig. 2). At two stan-
FIG. 1. Wear in the presence of each ester as a function of the percentage by volume of addi-
tive. Wear was determined by means of a high-frequency reciprocating rig. The identification
of vegetable oils and their sources is presented in the Materials and Methods section.
JAOCS, Vol. 78, no. 6 (2001)

582
D.C. DROWN ET AL.
FIG. 2. Wear at 0.1% volume addition of ester. Bold line at 670 µm represents the mean, dot-
ted lines are at ±1 standard deviation from the mean, and dashed lines are at ±2 standard de-
viations from the mean.
dard deviations, all samples were accounted for. This indicates tive, there was a 30% increase in lubricity for the ethyl esters
that at 95% confidence there was no statistical difference be- over the methyl esters. From the 1.0% additive (Fig. 4), the
tween any of the samples. Although the average lubricity was same trend could be seen, but the result was a 24% increase
slightly enhanced by adding 0.1% ester, there was no signifi- between the ethyl and methyl esters. The similarity of the per-
cant difference as to which type of ester sample was used.
centage increase between ethyl and methyl esters at both 0.5
Figure 3 shows the results of the lubricity testing for 0.5% and 1.0% indicated the same benefit of using ethyl esters over
ester additive where subsets 2 and 3, the methyl and ethyl es- methyl esters, regardless of the percentage of ester added to
ters excluding the castor oil samples, were evaluated. With the sample. This conclusion was supported by a Wilcoxon
the population separated into these subsets it was easy to ana- rank-sum (11,12) test, where for both 0.5 and 1.0% ester ad-
lyze the hypothesis of increased lubricity for the ethyl esters dition there was 100% confidence that the ethyl esters per-
as opposed to the methyl esters. In both cases the standard de- formed better than the methyl esters in the HFRR test.
viation of the subset was less than the standard deviation of
To determine if esters of castor oil were statistically differ-
the baseline measurements of 74.4 µm. The data within each ent, the following approach was taken. All data were plotted
subset had low statistical variation and could be represented as shown in Figure 3 for 0.5% volume additive and Figure 4
by the mean. In using this approach, for 0.5% volume addi- for 1.0% volume additive to show where the data bars for the
FIG. 3. Wear at 0.5% volume addition of ester. See Figure 2 for explanation of solid, dotted,
and dashed horizontal lines. Methyl and ethyl castor esters were evaluated separately from the
rest of the esters.
JAOCS, Vol. 78, no. 6 (2001)

ESTERS AS FUEL LUBRICITY ENHANCERS
583
FIG. 4. Wear at 1.0% volume addition of ester. See Figure 2 for explanation of solid, dotted,
and dashed horizontal lines. Methyl and ethyl castor esters were evaluated separately from the
rest of the esters.
castor oil ester were with respect to the other esters. From With the exception of castor, the esters showed increased vis-
Figure 3 one can easily see how the wear for methyl castor cosity for increasingly longer carbon chains. The ester from
lies far below the boundaries of the second standard devia- coconut oil had the shortest carbon chains, as well as the low-
tion line, which represents 95% confidence. Again the ethyl est kinematic viscosity. This trend continued up to the longest
castor sample lies outside of two standard deviations for the carbon chains in the Dwarf Essex HEAR.
ethyl esters. The same trend was found for the 1.0% volume
The conclusion that castor esters have improved lubricity
additive (Fig. 4). This proves that within 95% confidence the over the other esters was not fully understood. There was a dif-
esters of castor oil are statistically different and show im- ference in the fatty acid chemical structure, but the free glyc-
proved lubricity over the esters of the other oils tested.
erine content and kinematic viscosity measurements were also
The ethyl esters showed improved lubricity over the significantly different. Whether the improved lubricity of cas-
methyl esters of the same oil. This research also rejected the tor esters was due to the glycerine retained in the sample or to
hypothesis that the lubricity of the ester was influenced by the the attached OH group on the seventh carbon is unknown.
length of the carbon chains in the fatty acids of the oil. How-
It is important to determine what volume percentage of an
ever, there appeared to be a trend in the kinematic viscosity ester would be required to meet specific wear limits. The
of the ester with the chain length of the fatty acids in the ester. mean data were empirically fit to second-order polynomial
FIG. 5. Wear as a function of ester concentration. ꢀ, All of the methyl esters, taken together;
ꢁ, all of the ethyl esters, taken together; ꢂ, all of the castor esters, taken together.
JAOCS, Vol. 78, no. 6 (2001)

584
D.C. DROWN ET AL.
equations to produce the plot shown in Figure 5. The y in each
equation represents wear (µm) and x represents the ester con-
centration (%). To achieve the acceptable maximum wear scar
of 450 µm for the U.S. Army specification, about 0.21% of
castor ester, 0.39% ethyl ester, and 0.76% methyl ester must
be added to the JP-8 fuel. This figure can be used to determine
the economic balance between the type of alcohol ester used
and the costs to produce enough ester to achieve the desired
lubricity enhancement of the ester.
3. Lacey, P.I., and S.A. Howell, Fuel Lubricity Reviewed, SAE
Paper No. 982567 (October 1998).
4
. Nikanjam, M., Diesel Fuel Lubricity: On the Path to Specifica-
tions, SAE paper No. 1999-01-1479 (May 1999).
5
. Nikanjam, M., P.T. Henderson, T. Crosby, C. Gray, K. Meyer,
and N. Davenport, ISO Diesel Fuel Lubricity Round Robin Pro-
gram, SAE paper No. 952372 (October 1995).
6
. Bam, N., D.C. Drown, R.A. Korus, D.S. Hoffman, T.G. John-
son, and J.M. Washam, Method for Purifying Alcohol Esters,
U.S. Patent 5,424,467 (1995).
7
8
. HFRR Lubricity and Fretting Test System User Manual, PCS
Instruments, London, England (Sept. 1997).
. Annual Book of ASTM Standards 2000, American Society for
Testing and Materials, West Conshohocken, Pennsylvania,
2000, Section 5, Vols. 05.01, 05.02, 05.03, and 05.04.
ACKNOWLEDGMENTS
This research was supported in part by a research grant from U.S.
Army Tank-Automotive Research, Development, and Engineering
Center (TARDEC) Fuels and Lubricants Research Facility. Idaho
Transtech Inc. provided the high-erucic acid rapeseed oils. Lieu-
9. Shaver, B.D., R.M. Giannini, P.I. Lacey, and J. Erwin, Effects
of Water on Distillate Fuel Lubricity, SAE Paper No. 982568
(October 1998).
tenant Colonel Steve Zaat provided valuable advice on military re- 10. Bailer, J., P. Hodel, K. de Hueber, M. Mittelbach, C. Plank, and
quirements.
H. Schindlbauer, Handbook of Analytical Methods for Fatty
Acid Methyl Esters Used as Diesel Fuel Substitutes, Research
Institute for Chemistry and Technology of Petroleum Products,
University of Technology, Vienna, Austria, 1994, pp. 27–38.
REFERENCES
1
. Lemke, D., Betting on Biodiesel, Ag Innovation News, Agricul- 11. Devore, J., and N. Farnum, Applied Statistics for Engineers and
tural Utilization Research Institute, Vol. 9, (January 2000).
Website http://auri.org/news/ainjan00/diesel.htm.
Scientists, Duxbury Press-Brooks/Cole Publishing Co., Pacific
Grove, 1999, pp. 370–374.
2
. U.S. Air Force, MIL-DTL-83133E Turbine Fuels, Aviation, 12. Box, G.E.P., W.G. Hunter, and J.S. Hunter, Statistics for Exper-
Kerosene Types, NATO F-34 (JP-8), NATO F-35, and JP-8 +
00, ASC/ENSI Bldg., Wright-Patterson Air Force Base, Day-
ton, 1999.
imenters, John Wiley & Sons, New York, 1978, pp. 80–82.
1
[Received July 10, 2000; accepted March 16, 2001]
JAOCS, Vol. 78, no. 6 (2001)