Note: Descriptions are shown in the official language in which they were submitted.
CA 02607478 2007-10-23
FUEL COMPOSITION
BACKGROUND INFORMATION
[0001] Diesel engines continue to find wide use in trucks, ships, trains, and
the like.
Commercially acceptable diesel fuel must be capable of performing over a range
of
climatic conditions and, accordingly, must be able to be used at temperatures
down to
0 C and preferably as low as at least -10 C.
[0002] Diesel engine exhaust often includes particulates, CO, and various
nitrogen
oxide (NOX) species. Over the years, many attempts to address the
environmental
concerns associated with exhaust fumes from diesel engines have focused on
lower
alcohols such as methanol and ethanol. A blend of 15% (by vol.) ethanol and
85% (by
vol.) diesel oil has been found to improve the combustion byproducts emitted
in the
engine exhaust (generally believed to be due to the increased oxygen content
of the fuel)
without requiring modifications to existing diesel engines; in recent years,
the amount
of ethanol has been increased at times to 20% (by vol.). However, lower
alcohols typi-
cally are immiscible with diesel oil and tend to separate over time, so the
components
often are stored separately and mixed just prior to use.
[0003] Fuel blends of the type just described often are referred to as "E-
diesel" (or
some similar variant). Although E-diesel typically generate fewer
objectionable
combustion byproducts than neat diesel oil, it produces less energy when
combusted
and still employs petroleum-derived diesel oil for at least 80% of its volume.
Sustain-
ability and sourcing concerns regarding fossil fuels have grown significantly
over the
past decade. In turn, this has increased interest in fuels prepared from
sources other
than petroleum has grown significantly over the past decade or so.
1
CA 02607478 2007-10-23
[0004] Much of this interest has focused on so-called biofuels, which are the
transesterification products of any of a variety of animal fats and vegetable
oils. The
major components of oils and fats are fatty acid triglycerides, molecules in
which three
long chain fatty acids are ester linked to a glycerol radical. When the oil or
fat is
exposed to an alcohol (typically methanol) under appropriate catalytic
conditions, the
fatty acids can cleave from the glycerol radical and react with the alcohol to
form fatty
acid esters. The transesterification reaction significantly reduces the
viscosity of the oil.
[0005] The triglyceride transesterification reaction has been much studied;
for more
information on the manufacture and properties of such biofuels, the interested
reader is
directed to any of a variety of overviews such as, e.g., M. Graboski et al.,
"Combustion
of Fat and Vegetable Oil Derived Fuels in Diesel Engines," Prog. Energy
Combust. Sci.,
vol. 24, pp. 125-64 (1998, Elsevier).
[0006] Biofuels can be used neat but, more commonly, small proportions are
blended into petroleum-derived diesel (hereinafter "petrodiesel"). Blends
of.biofuel
and petrodiesel often are referred to as B-diesel or, more commonly, with a
number
following the B to indicate the percentage of petrodiesel replaced with
biodiesel (e.g.,
B20 diesel indicating of blend of 80% petrodiesel and 20% biodiesel).
[o007] Some have attempted to combine alcohol and biofuel in a single
composition;
see, e.g., U.S. Patent No. 6,129,773 and U.S. Patent Appl. Publ. No. US
2003/0126790 Al.
[0008] Fuel compositions that include components derived solely from renewable
starting materials appear to be of continued and growing interest. Ideally,
such a
biofuel composition could be used in essentially the same climatic conditions
as, and
have emission characteristics that are better than, petrodiesel. Additionally,
such
compositions also preferably would produce nearly as much energy as an
equivalent
volume of petrodiesel while resulting in similar or better engine wear
characteristics.
2
CA 02607478 2007-10-23
SUMMARY
[0oo9] In one aspect is provided a fuel composition that includes ethanol and
a C2-C6
ester of one or more long chain fatty acids. The composition generally
includes at least
about 2.5% (by vol.), typically from about 5 to about 10% (by vol.), of a
lower alkyl
monool such as ethanol and a complementary amount of long chain fatty acid
ester(s);
all other components typically are present in no more than trace amounts.
[0010] In some embodiments, at least 99.99% (by wt.) or even 99.999% (by wt.)
of the
composition can constitute just C, H and 0 atoms; in these and other
embodiments, the
composition can be essentially free of at least one of, and preferably both
of, sulfur and
nitrogen atoms.
[0011] The composition typically includes at least about 0.2% (by vol.),
commonly at
least about 0.25% (by vol.), and occasionally at least about 0.3% (by vol.)
water. In some
embodiments, the fuel composition can include at least about 0.5% (by vol.)
water.
Water in a fuel composition typically is considered something to be avoided if
at all
possible, yet the presence of up to about 1 % (by vol.) has not proven to be
particularly
deleterious to the efficacy of the present fuel composition.
[0012] In some embodiments, the fuel composition can have an acidic pH, at
times
as low as, e.g., 4.5, but more commonly from about 6.0 to about 6.8.
[0013] The composition typically has a kinematic viscosity of from about 3.5
to about
4.0 mm2/s (i.e., cSt), commonly about 3.7 0.2 mm2/s. Even in the absence of
flow
improving additives, the composition can have a cloud point of at least as low
as about
-5 C and a pour point of at least as low as about -15 C.
[0014] In another aspect is provided a method for synthesizing and refining a
fuel
composition. The method includes providing a liquid that contains at least one
C2-C6
ester of one or more long chain fatty acids and adjusting the pH of the liquid
to less
than 7Ø The C2-C6 ester(s) can be provided by transesterification of a
triglyceride-
containing composition using a C2-C6 monool and, in such cases, the pH
adjustment
3
CA 02607478 2007-10-23
results in separation of a salt and glycerine-related byproducts (e.g.,
glycerol) from the
liquid phase, thereby resulting in a raw fuel composition. The raw fuel
composition
then is treated so as to remove particulates having an average diameter
greater than -1
m and to ensure that the concentration of Group I metal ions is less than
about 50 parts
per million (ppm), preferably even 5-10 times lower than this number.
[0015] The transesterification process typically is performed in the presence
of a
stoichiometric excess of one or more C2-C6 alcohols. The alcohol(s) typically
are present
in from -25 to -200% excess (by vol.), preferably from -50 to -150% excess,
more
preferably from -70 to -130% excess, and most preferably from -80 to -120%
excess.
[0016] The pH of the liquid, as well as the raw fuel composition and the fuel
composition that ultimately results from the refining process (at least
initially), typically
is between about 4.5 and about 6.8, preferably of from about 5.0 to about
6.75, more
preferably of from about 5.5 to about 6.7, and most preferably from about 6.0
to about
6.65. Refining the fuel composition in a slightly acidic condition can provide
storage
stability benefits.
[0017] In certain embodiments, the pH adjustments is accomplished by adding a
strong acid to the liquid. In some of these embodiments, the acid is
halogenated (e.g.,
HCl), and the salt removed from the liquid is KX where X is a halogen atom
(e.g., Cl).
[0018] In some embodiments, the raw fuel treatment can be accomplished by
passing the raw fuel composition through a series of filters, optionally of
progressively
smaller pore sizes.
[0019] The process can provide a fuel composition that has a kinematic
viscosity of
from about 3.5 to about 4.0 mm2/s, most commonly of no more than about 3.8
mm2/s.
[0020] The process also can provide a fuel composition that remains fluid even
in the
absence of flow improving additives. Specifically, the composition can have a
cloud
point at least as low as -5 C and a pour point at least as low as -15 C. The
cetane
4
CA 02607478 2007-10-23
number of a refined fuel composition typically is greater than about 45 and
can be
significantly (e.g., 5-15%) higher.
[0021] Advantageously, all steps of this process can be performed without
utilizing
external sources of heat, i.e., at or near ambient temperatures. Additionally,
the process
can be performed without the addition of significant amounts of water, thereby
elimi-
nating the need to collect and treat (typically very caustic) waste water.
[0o22] In yet another aspect is provided a method of powering a vehicle. The
method involves introducing a fuel that includes the aforesaid fuel
composition into a
diesel engine and allowing the engine to combust the fuel. Advantageously, the
afore-
said fuel composition can result in simultaneous reductions in opacity, CO,
SO2, and
NOX in the exhaust compared to the exhaust of the same engine combusting
petrodiesel.
Simultaneously, the amount of 02 in the same exhaust increases relative to
that seen
when petrodiesel is combusted. In many cases, the aforesaid composition also
greatly
reduces the amount of and, in some instances, removes carbonaceous deposits
(often
referred to as "coking") on metal parts of the engine. With the probable
exception of
the detergency effect and lower exhaust temperatures, the other
characteristics might be
more evident in a two-cycle diesel engine than in a four-stroke diesel engine.
(0023] Other aspects of the invention are set forth in the more detailed
description
which follows.
DETAILED DESCRIPTION
[0024] The process of making a fuel composition, more specifically a process
that
involves synthesizing and refining a fuel composition, is described first.
[0025] Like other biofuel manufacturing processes, the synthesis portion of
the
process involves a transesterification reaction. However, certain aspects of
the reaction
differ from that which has become conventional in the manufacture of biofuels.
CA 02607478 2007-10-23
[0026] Almost any triglyceride can be transesterified with an alcohol. For
example,
as described above, some are using the filtered waste oil from fast food
restaurants as a
low-cost reactant. However, the present process gives preference to highly
pure
vegetable oils, more preferably to food-grade materials, examples of which
include
corn, linseed, peanut and soybean oils.
[00271 For the convenience of the reader, the following tables provide
approximate
percentages of various saturated and unsaturated fatty acid components of a
variety of
fats and oils. The first is from page 130 of the Graboski et al. article
mentioned
previously, while the second is from Appendix A of C.L. Peterson et al.,
"Performance
and Durability Testing of Diesel Engines Using Ethyl and Methyl Ester Fuels,"
National
Biodiesel Board Report for 1995, received for publication on February 27,
1996. The
former provides data for raw materials (i.e., ones that have not undergone
transesterifi-
cation), while the latter provides data for esterified fatty acids in biofuels
made from
such materials.
6
CA 02607478 2007-10-23
Table la: Weight percentages of fatty acids in fats and oils
Carbon number
Saturated acids Mono-unsaturated acids Di Tri
8 10 12 14 16 18 >18 <16 16 18 >18
Beef tallow - - 0.2 2-3 25-30 21-26 0.4-1 0.5 2-3 39-42 0.3 2 -
Butter 1-2 2-3 1-4 8-13 25-32 8-13 0.4-2 1-2 2-5 22-29 0.2-1.5 3 -
Coconut 5-9 4-10 44-51 13-18 7-10 1-4 - - - 5-8 - 1-3 -
Cod liver - - - 2-6 7-14 0-1 - 0-2 10-20 25-31 35-52 - -
Corn - - - 0-2 8-10 1-4 - - 1-2 30-50 0-2 34-56 -
Cottonseed - - - 0-3 17-23 1-3 - - - 23-41 2-3 34-55 -
Lard - - - 1 25-30 12-16 - 0.2 2-5 4-51 2-3 4-22 -
Linseed - - - 0.2 5-9 0-1 - - - 9-29 - 8-29 45-67
Palm - - - 1-6 32-47 1-6 - - - 40-52 - 2-11 -
Palm 2-4 3-7 45-52 14-19 6-9 1-3 1-2 - 0-1 10-18 - 1-2 -
kernel
Peanut - - - 0.5 6-11 3-6 5-10 - 1-2 39-66 - 17-38 -
Rapeseed - - - - 2-5 1-2 0.9 - 0.2 10-15 50-60 10-20 5-10
Safflower - - - - 5.2 2.2 - - - 76.3 - 16.2 -
Soybean - - - 0.3 7-11 3-6 5-10 - 0-1 22-34 - 50-60 2-10
Sunflower - - - - 6 4.2 1.4 - - 18.7 - 69.3 0.3
Tung - - - - - - - - - 4-13 - 8-15 72-88
Table 1b: Weight percentages of fatty acids in fats and oils
Carbon number
Saturated acids Mono-unsaturated acids Di Tri
14 16 18 >18 18 20 22
Beef tallow 3 23-24 18 0 38-39 0 0 0 0
Canola 0 4 2-3 0.5 65 1-2 0.1-0.2 17-18 7-8
Rapeseed 0 2-3 1 0.5-1 12-13 7-8 49-50 12 7-8
Soybean 0 10 3-4 0-0.5 19 0.2 0 55-56 10-11
(The numbers in Table lb do not add to 100% in all cases because some
materials
contained other, non-analyzed constituent fatty acids.)
[0028] A preferred starting material is a food-grade vegetable oil selected
from the
foregoing list. Particularly preferred are those that are refined, bleached,
and deodor-
ized (RBD); such materials are available from a variety of commercial sources
including,
7
CA 02607478 2007-10-23
for example, ConAgra, Bunge Ltd., ADM and the like. The remainder of this
description is based on an RBD soybean oil although, in view of the advantages
found
with this material and its composition relative to that of other oils, the
ordinarily skilled
artisan should be able to identify other sources of long chain fatty acids
that can provide
similar advantages and/or advantages in different end-use conditions.
(0029] The fatty acid source(s) can be provided in or to a reaction vessel.
While the
art teaches that most any type of vessel can be used, preparation of a high
quality
biodiesel-type fuel composition is facilitated by use of a vessel made from a
clean,
essentially non-reactive material. Preference here is given to materials such
as steel,
stainless steel, glass-lined metals, and the like.
[0030] The other reagent in transesterification reactions is an alcohol. By
far, the vast
majority of biofuels being made employ methanol as a reagent. In fact, the
European
Biodiesel Board mandates the use of methanol where a producer wishes to have
its
product certified as a biodiesel fuel; see, e.g., EN14214. However, distinct
advantages
have been found when higher alcohols, specifically one or more C2-C6 alcohols,
are
employed. Non-limiting examples of these benefits include improved miscibility
with
petrodiesel and better low temperature performance as exemplified by, e.g.,
cloud point
and cold pour point.
[0031] Each alcohol employed preferably is aliphatic, and more preferably has
the
general formula CnH2n+1OH where 2< n< 6. As indicated by the formula, the use
of
monools is preferred so as to avoid the formation of longer chain diesters.
Preferred
alcohols include ethanol, 1-propanol, 2-propanol, and 1-butanol, with absolute
ethanol
(denatured by means of, e.g., gasoline) being particularly preferred.
[0032] Where the emission characteristics of the resulting fuel are of
concern, the
alcohol preferably is free of heteroatoms that can be involved in the
formation of
undesirable species; non-limiting examples of such heteroatoms include N, S,
and P.
8
CA 02607478 2007-10-23
[0033] Reaction between the alcohol(s) and triglyceride(s) is catalyzed by
both acids
and bases, with a strong base being more commonly used. Typically, no more
than
-1 % (by wt.) catalyst is required, based on the triglyceride starting
material(s).
[0034] Conveniently, the catalyst can be delivered in some or all of the
alcohol(s).
This can be done by dissolving a strong base, e.g., KOH, in the alcohol(s)
prior to
delivery of the alcohol(s) to the reaction vessel. Conveniently, this can be
done in a
relatively short amount of time (less than an hour) through simple mixing at
ambient or
slightly elevated temperatures. If desired, the catalyst solution can be added
stepwise,
i.e., through the addition of sequential aliquots followed by mixing or
agitation.
[0035] An excess of alcohol preferably is introduced to the reaction vessel.
As noted
previously, the fuel composition contains both C2-C6 esters of long chain
fatty acid(s)
and C2-C6 alcohol(s). Rather than delivering a stoichiometric amount of
alcohol to the
reaction vessel and then later blending the resulting transesterified
product(s) with
additional alcohol, certain processing and performance advantages might be
obtained
by introducing the alcohol component(s) during the transesterification step.
In practice,
a significant excess of alcohol can be added to the reaction vessel; the
amount typically
ranges from -25 to -200%, preferably from -50 to -150%, more preferably from -
70 to
-130%, even more preferably from -80 to -120%, and most preferably at least -
100%
excess (all by volume).
[0036] Despite the aforementioned preference for using an excess of alcohol
during
the manufacture and refining steps, at least some of the benefits might be
able to be
achieved by post-manufacture addition of alcohol during or after refining.
[0037] To a certain extent, the stoichiometric excess of alcohol can be varied
somewhat so as to create fuel compositions with slightly different properties.
In this
manner, one can create seasonal blends of fuel. For example, during cold
weather
seasons, the excess of alcohol can be increased so as to provide a fuel
composition
having a lower viscosity than a similar composition intended for summertime
use.
9
CA 02607478 2007-10-23
[0038] The process has been described as involving introduction of the basic
alcohol
solution to the oil, although this is not to be considered limiting. Where
this is done,
however, introducing the alcohol solution near the bottom of the tank and
allowing it to
partition upwardly can provide some reaction advantages, particularly where
the
reaction mixture is not agitated, stirred, etc., and where the
transesterification reaction
is permitted to proceed at or near ambient temperatures. Where conventional
agitation
techniques (e.g., a circulating pump) are employed, the transesterification
reaction
generally proceeds to completion in a few hours at 20 -25 C. The
transesterification
reaction can be expedited with relatively gentle heating although this
certainly is not
required.
[0039] The second product of the transesterification reaction is glycerol and
other
glycerine byproducts. Depending on the identity, nature, and purity of the
reagents,
the glycerol in the reaction vessel can separate on its own, can partially
separate, or
separate only minimally. For example, where high purity reagents are used,
glycerol
(and derivatives) typically separates from the alkyl ester / ethanol phase in
a few hours
(e.g., 4-12 hours) after agitation is ceased. However, where a less refined
starting
material is employed, addition of more alcohol (e.g., up to -20% additional
alcohol)
and/ or additional time can be needed to achieve the desired level of
separation. The
alcohol added for separation purposes need not be the same as that used in the
reaction
and, in fact, methanol can be used for this purpose, if desired.
[0040] The glycerol phase is heavier than the alkyl ester / alcohol phase and
thus
separates to the bottom of the reaction vessel. Where the reaction vessel
includes an
egress at or near its bottom, the glycerol layer can be drained from the
reaction vessel.
Typically, the glycerol phase constitutes from -10-20% of the triglyceride
reactant.
Based on the excess amount of alcohol used during the transesterification
reaction, the
glycerol layer is presumed to contain a not insignificant amount of alcohol.
Processing
and/ or disposal techniques for glycerol are known.
CA 02607478 2007-10-23
[0041] Assuming that the material used as the basic catalyst contains a Group
I or II
metal atom (e.g., KOH or Ca(OH)2), the majority of the Group I or II metal
ions partition
into the glycerol phase. Removing the glycerol phase thereby conveniently
removes the
majority of Group I or II metal ions present in the reaction vessel. The
remaining alkyl
ester/ alcohol phase typically contains on the order of 500 to 1000 ppm of
such ions,
more commonly on the order of$00 to 800 ppm.
[0042] The alkyl ester/ alcohol blend can be further treated in the reaction
vessel or,
more commonly, transferred to one or- more additional vessels for further
refining. At
this point, the blend typically has a kinematic viscosity of -6.2 to -6.8
mm2/s.
[0043] Organizations such as the National Biodiesel Board (NBB) suggest
techniques
in which the raw fuel (which, assuming stoichiometric amounts of triglyceride
and
methanol, will be extremely caustic) is washed repeatedly with water. This
repeated
washing uses large amounts of another natural resource (i.e., water) and
results in huge
volumes of moderately to extremely caustic waste water which, in turn, must be
treated
before and/or after disposal.
[0044] Conversely, the present process gives preference to techniques that do
not
employ water per se. For example, the raw fuel composition can be passed by or
through a cationic exchange resin which replaces Group I or II metal ions with
H atoms.
Such resins are widely known and available from a variety of commercial
sources
including, e.g., Rohm and Haas Co. (Philadelphia, Pennsylvania).
[0045] Alternatively, relatively small amounts of a strong acid can be used to
neutralize the alkyl ester/ alcohol blend. Halogen-containing strong acids
(e.g.,
concentrated HCI) are preferred over other strong acids that contain the types
of
heteroatoms mentioned previously, e.g., H2SO4, HNO3r etc. While the ordinarily
skilled
artisan can perform the stoichiometric calculations needed to determine the
amount of a
given acid needed to neutralize a given raw fuel composition, by way of non-
limiting
example, each liter of concentrated HCl can treat over 500 L of raw fuel
composition.
11
CA 02607478 2007-10-23
[004q Such refining techniques result in removal of Group I or II metal ions.
Where
acid is added to the blend, the ions are removed as salt along with a very
significant
portion of any glycerol-type byproduct that did not phase separate previously.
The
latter is evidenced by a fairly significant reduction in kinematic viscosity.
As
mentioned previously, a raw fuel composition typically has a kinematic
viscosity at
40 C of -6.5 mm2/ s; conversely, the same composition having undergone the
type of
acidulation refining just described typically has a kinematic viscosity at 40
C of from
-4.0 to -5.5 mm2/s. This is believed to be somewhat better than that which can
be
attained through simple water washing and/or use of ion exchange resins. For
example, in the aforementioned Peterson et al. article, the various ethyl
ester biofuels
had kinematic viscosities of from 4.5 mm2/s (soybean oil) to 6.2 mm2/s
(rapeseed oil).
[0047] Regardless of which refining technique is used, the processed fuel
composition (i.e., purified alkyl ester/ alcohol blend) preferably is rendered
very
slightly acidic, whereas generally accepted techniques which argue for a
refined fuel
that is essentially neutral (pH = 7.0). At this point in the refining process,
the alkyl
ester/ alcohol blend can register a pH of as low as 4.0 to 4.5, typically
between about 4.5
and about 6.9; more commonly, the pH of the blend will be in one or more of
the
following ranges: about 5.0 to about 6.8, about 5.5 to about 6.75, about 5.75
to about 6.7,
about 5.9 to about 6.7, about 6.0 to about 6.75, about 6.0 to about 6.7, about
6.1 to about
6.6, about 6.1 to about 6.7, and about 6.4 0.2.
[0048] (For fuel compositions, particularly biofuels, acid numbers commonly
are
reported. This is a measure of free, i.e., non-esterified fatty acids and is
given in terms
of mg KOH necessary to neutralize those fatty acids. When a strong protic acid
is
utilized in the refining of the present fuel composition, this acid number is
believed to
be far less meaningful. For at least this reason, a conventional pH meter was
employed
to obtain pH measurements, and these are the numbers used herein.)
12
CA 02607478 2007-10-23
[0049] By providing a fuel composition with a slightly acidic pH, storage and
handling advantages have been observed without any noticeable deleterious
effects on
performance. Specifically, fuel compositions refined so as to have an
essentially neutral
pH at this step have been observed to form a flocculent or precipitate when
stored for
extended periods and/or, particularly, when exposed to air. Conversely, fuel
compo-
sitions with a slightly acidic pH have not been observed to suffer from this
tendency.
[0050] At this point of the refining process, the fuel composition includes
primarily
alkyl esters of long chain fatty acids and alcohol; most other materials are
present in
essentially just trace amounts. However, the fuel composition generally
includes from
about 0.2 to about 0.9% (by vol.), commonly from about 0.25 to about 0.75% (by
vol.),
and at times from about 0.3 to about 0.6% (by vol.) water. This amount of
entrained or
dispersed water is not significantly reduced even when the fuel composition is
filtered,
as discussed in more detail below; nevertheless, the presence of such water
has not been
found to have significant deleterious effects on combustion of the fuel
composition and
might even provide certain benefits (e.g., reduced combustion temperatures).
[0051] Additionally, at this stage of the refining process, the amount of
Group I or II
metal ions typically is reduced to no more than about 50 ppm, preferably no
more than
about 25 ppm, more preferably no more than about 10 ppm, even more preferably
no
more than about 5 ppm, and most preferably no more than about 4 ppm.
[0052] The fuel composition can include up to about 50% (by vol.) alcohol
relative to
the overall volume. The fuel composition generally can include from about 2 to
about
40% (by vol.) alcohol, although from about 3 to about 30% (by vol.) is more
common
and from about 4 to about 20% (by vol.) is most common. Where high purity
reagents
are employed (as described above), a refined fuel composition made according
to this
process typically includes from -5 to -15% (by vol.), commonly from -5.5 to -
10% (by
vol.), and most commonly from -6 to -8% (by vol.) alcohol; preferably one or
more
C2-C4 monools such as ethanol and/or 1-butanol.
13
CA 02607478 2007-10-23
[0053] At this point, several options are available. For example, the refined
blend
can be used as is or can be used after addition and thorough mixing of one or
more
conditioners, stabilizers, or other additives (e.g., kerosene). However,
certain further
advantages can be obtained by additional refining of the fuel composition,
optionally
containing additives of the types of just discussed.
[0054] An additional refining technique that has been found advantageous in
certain
circumstances is filtering. Specifically, the fuel composition can be passed
through one
or more filters, optionally of progressively smaller pore size, so as to
remove suspended
contaminants. Commercial filtration devices are available from a variety of
sources
including, e.g., Donaldson Co., Inc. (Minneapolis, Minnesota), Central
Illinois Manufac-
turing Co. (Bement, Illinois), Harvard Corporation (Evansville, Wisconsin),
and Wix
Filtration Products (Gastonia, N. Carolina). Using a pump to pressurize the
system to
-130 to -140 kPa can provide a processing rate on the order of -550 to -700
mL/s.
[0055] Performing such filtration on raw fuel compositions has not been found
to
provide significant advantages, at least on a consistent basis. However, when
filtration
is done on a fuel composition that has been treated by the aforementioned
acidulation
technique, fuel compositions having kinematic viscosities at 40 C on the order
of no
more than -4.0 mm2/ s, of no more than -3.9 mm2/ s, of no more than -3.8 mm2/
s, and
even of no more than -3.7 mm2/s can be obtained. This technique is believed to
be
capable of providing an ethyl ester of soy oil/ ethanol fuel composition with
a kine-
matic viscosity at 40 C on the order of -3.6 mm2/s, -3.5 mm2/s, or even lower.
These
viscosity values are in contrast to those reported for prior art ethyl ester
biofuels; see,
e.g., the Peterson et al. article data mentioned above as well as the Graboski
et al. article
(from 4.4 to 5.9 mm2/s). Additionally, the presence of free alcohol(s) in the
fuel compo-
sition can explain no more than about half of the viscosity reduction seen in
the present
fuel composition. An explanation for the remainder of this reduction is not
fully under-
14
CA 02607478 2007-10-23
stood but might result from the acidulation step making one or more of the
undesired
byproducts more susceptible to removal by further refining steps such as,
e.g., filtration.
[0056] Thus, a fuel composition that has been both acidulated and filtered can
have a
kinematic viscosity that is on the order of 40% less than that of the raw fuel
from which
it has been refined. Because petrodiesel generally is expected to have a
kinematic
viscosity of no more than 4.1 mm2/s when measured at 40 C in accordance with
ASTM
D975, providing a biofuel with similar viscosity characteristics can be
advantageous
with respect to both commercial acceptance and in-use performance. The
ordinarily
skilled artisan understands the desirability of having a biofuel composition
that has
storage and performance characteristics that are as similar as possible to
ubiquitous
petrodiesel. For example, published reports including the aforementioned
Peterson et
al. article indicate that coking of fuel injectors might correlate directly to
fuel viscosity.
Coking also has been surmised to be due to impurities in the biofuel; see
again, e.g., the
Peterson et al. article. These two theories might be related: refining that
leads to fewer
impurities likewise might result in a reduction in viscosity.
[0057] The refining process just described is believed to provide significant
benefits
over those commonly employed in the manufacture of biodiesel fuel. For
example, this
process does not require the use of large volumes of water to wash Group I
ions out of
the raw fuel composition; in turn, this reduces the amount of water that must
be used
and treated prior to disposal. Additionally, because excess alcohol in the
refined fuel
composition is desirable, this process does not require the use of time- and
energy-
intensive distillation techniques. Thus, in addition to using essentially only
natural
starting materials, the process requires the input of very little energy to
make and refine
a fuel composition.
[0058] Once fully refined, the fuel composition can be stored without a need
for
significant treatments or precautions. As mentioned previously, by refining
the fuel
composition in a slightly acidic form, better storage and handling performance
can be
CA 02607478 2007-10-23
achieved. However, whether the fuel composition must be maintained in acidic
form
once refining is completed has not been determined. In other words, the pH of
the
refined fuel composition might be able to be adjusted upwardly so that fuel
composi-
tion has an essentially neutral pH prior to use without negatively affecting
the viscosity
and storage stability of the fuel composition once the refining process is
complete.
[0059] Advantageously, this process can result in a fuel composition having a
cloud
point (as measured in accordance with ASTM D2500) of at least as low as about -
2 C,
-3 C, -4 C, -5 C, -6 C, -7 C, or even lower, and a pour point (as measured in
accordance
with ASTM D97) of at least as low as -10 C, -12.5 C, -15 C, -17.5 C, -20 C, or
even
lower. For example, a fuel composition made from ethanol and RBD soybean oil
according to the foregoing process, treated with a small amount of diesel
conditioner,
was found to be in useful condition even after sitting outside overnight in
air temper-
atures that fell to at least -20 C; commercially available biodiesel fuels,
even those with
significant amounts of conditioners and other additives, are not believed
capable of
achieving this type of low temperature performance. For example, the Peterson
et al.
article reports that ethyl esters of fatty acids have pour points of from -10
(ethyl ester of
rapeseed oil) to 12 C (ethyl ester of beef tallow). By way of a more direct
comparison,
that same article indicates that an ethyl ester of soybean oil has a pour
point of -3 C;
accordingly, the present.process appears to be capable of providing fuel
compositions
with pour points that are at least 5 -20 C lower than those of standard ethyl
ester
biofuels.
[0060] A fully refined fuel composition can be used as is or, depending on the
end
use application, diluted with an appropriate amount of petrodiesel. For
example, some
fueling locations have created a 50:50 blend of bio- and petrodiesel and then
used this
blend as a masterbatch for providing further diluted blends. To date, no
significant
miscibility issues have been reported, even with the so-called masterbatch
blends.
16
CA 02607478 2007-10-23
[0061] As suggested previously, a fuel composition according to the present
invention generally includes a lower alkyl monool (e.g., ethanol) and a C2-C6
ester of
one or more long chain fatty acids. The composition generally includes from
about 5 to
about 10% C2-C4 alcohol(s), preferably ethanol, and a complementary amount of
long
chain fatty acid ester(s); all other components typically are present in no
more than
trace amounts. In some embodiments, at least 99.99% (by wt.) or even 99.999%
(by wt.)
of the composition can constitute just C, H and 0 atoms; in these and other
embodi-
ments, the composition can be essentially free of at least one of, and
preferably both of,
sulfur and nitrogen atoms.
[0062] The composition generally includes water, typically in an amount of
from
about 0.2 to about 0.5 % (by vol.) and amounts of as much as 0.8% (by vol.) or
more are
believed possible in certain circumstances.
[0063] For reasons already discussed, the fuel composition preferably has a
slightly
acidic pH (at least during refining) and a kinematic viscosity at 40 C of
about 3.7 0.2
mm2/s. Even in the absence of flow improving additives, the composition can
have a
cloud point of at least as low as about -5 C and a pour point of at least as
low as -15 C.
Each of these properties can be achieved in isolation or, in some embodiments,
in
combination.
[0064] A biofuel composition of this type can be used neat and, in some
circumstances, can provide significant emission advantages over neat
petrodiesel or a
blend of petrodiesel and biofuel.
[0065] For example, a fuel composition including ethanol and an ethyl ester
trans-
esterification product of RBD soybean oil was tested in a short rail line
locomotive
(EMD model 16-645BC) employing a two-cycle, V-16, roots blown, non-
turbocharged
engine; each cylinder had a displacement of -10.5 L (645 cubic inches),
resulting in a
total displacement of nearly 170 L. Prior to testing, the locomotive was
provided with
-280 L (75 gallons) of neat biofuel composition and allowed to warm up,
thereby
17
CA 02607478 2007-10-23
flushing any remaining petrodiesel from the engine. Thereafter, -190 L (50
gallons) of
each of the following fuels were tested sequentially: neat biofuel, a 50:50
blend of
biofuel composition and railroad off-road #2 petrodiesel, and neat
petrodiesel.
Emissions testing was performed with the engine under a steady state load
created by
connecting the locomotive's diesel powered DC generator to a loading grid
designed to
convert electrical power to heat.
[0066] Percent smoke opacity was measured continuously using a WagerTM 7500
smoke meter (Robert H. Wager Co., Inc.; Rural Hall, North Carolina) clamped on
a 5 cm
(2 inch) sampling elbow tube placed inside one of the exhaust stacks of the
locomotive.
[0067] Emission gases were measured using a testoTM 350XL portable gas
analyzer
(testo, Inc.; Flanders, New Jersey) set to measure exhaust 02, CO, SO2, total
hydro-
carbons, NO, NO2r and combined NOx levels. (Total hydrocarbon data was not
collected due to a sampling issue.) The analyzer came equipped with a model
450
control unit and attached sampling probe. The gas analyzer probe was
maintained in a
constant sampling position by a bracket that held the probe approximately 5 cm
(2 inch)
into the center of the exhaust stack.
[0068] Data collection began once the opacity readings stabilized after
introduction
of each test fuel. After this initial recording, readings were continuously
charted in 1-2
minute intervals.
[0069] The results of this testing summarized below in Table 2 are averages of
five of
these values.
18
CA 02607478 2007-10-23
Table 2: Exhaust Characteristics, Two-Cycle Diesel Engine
100% petrodiesel 50:50 blend 100% biodiesel
Temp. ( C) 419 404 370
Opacity (%) 15 4 2
[S02] (ppm) 89 28 0
[02] (vol. %) 12.7 13.2 13.9
[CO] (ppm) 257 120 60
[NOX] (ppm) 1225 1114 1047
[0070] The data from Table 2 indicate many interesting characteristics. For
example,
because the fuel composition according to the present invention was
synthesized and
refined in a manner that avoided the introduction of S atoms, the 100%
reduction
relative to #2 diesel oil (petrodiesel) might be explainable, although it
still is better than
results reported previously (e.g., the Peterson et al. article). Additionally,
because the
inventive fuel composition resulted in a lower exhaust temperature, the
reduction in
NOX emissions also might be readily explainable (because NOX formation is
known to
increase as combustion temperatures increase). Further, the higher oxygen
content of
the inventive fuel composition simultaneously increased the 02 content and
greatly
reduced (i.e., more than 75%) the CO concentration of the exhaust.
[0071] The foregoing results contrast with those of a previously published
study
comparing off-road #2 petrodiesel (one meeting current EPA regulations and
another
meeting stricter California regulatory standards) with B20 diesel, i.e., an
80:20 blend of
petrodiesel and a methyl ester derivative of soybean oil. See S.G. Fritz,
"Evaluation of
Biodiesel Fuel in an EMD GP38-2 Locomotive," National Renewable Energy
Laboratory
Subcontractor Report dated May 2004 (available from the U.S. Dept. of
Commerce).
[0072] That study was performed on a road-switcher locomotive equipped with an
EMD 16-645-E diesel engine having the characteristics set forth below in Table
3.
19
CA 02607478 2007-10-23
Table 3: Specifications, Engine Used in Comparative Test
engine type 2-stroke, uniflow scavenged diesel
aspiration roots blown, non-turbocharged
bore (mm) 230
stroke (mm.) 254
number of cylinders 16
displacement (L), each cylinder 10.6
total displacement (L) 169
compression ratio 16:1
rated power (kW) 1491
rated speed (rpm) 900
[0073] That exhaust emissions study results are provided below in Table 4,
where
the following acronyms and abbreviations are used:
EPA - U.S. Environmental Protection Agency
EPA Cert Baseline - diesel fuel meeting EPA specifications for locomotive
emissions (see 40 C.F.R. 92.113)
CARB - a 50:50 blend of two commercially available fuels
meeting California Air Resources Board specifications
B20 - a 20:80 blend of G-3000TM biofuel (Griffin Industries, Inc.;
Cold Spring, Kentucky) and EPA Cert Baseline
C20 - a 20:80 blend of G-3000TM biofuel and CARB
C.O.V. - coefficient of variation
HC - total hydrocarbons
PM - particulate matter
AAR - Association of American Railroads
AAR Corr. BFSC - AAR-corrected brake-specific fuel consumption
CA 02607478 2007-10-23
Table 4: Results from Comparative Emissions Testing
EPA Line-Haul Duty Cycle EPA Slvitch Duty Cycle
HC CO NOx PM AAR Corr. HC CO NOx PM AAR Con-
gRip-hr glhp-hr g/hp-hr gRip-hr BSFC /hp-hr glhp-hr glhp-hr glhp-hr BSFC
Iblhp-hr Iblhp-hr
EPA Cert Baseline #1/3 0.71 5.9 11.9 0.47 0.433 0.87 2.4 12.7 0.37 0.464
EPA Cert Baseline #2/3 0.62 5.1 12.3 0.49 0.435 0.80 2.2 12.9 0.42 0.474
EPA Cert Baseline #3/3 0.58 5.1 12.9 0.44 0.434 0.78 2.1 12.9 0.36 0.460
Average 0.64 5.4 12.4 0.46 0.434 0.82 262 12.8 0.38 0.466
C.o.v. 10% 9% 4% 5% 0%. 6% 7% 1% 8% 2%
CARB #1/3 0.63 4.6 12.2 0.48 0.431 0.78 1.8 12.7 0.34 0.467
CARB #213 0.62 4.7 12.0 0.46 0.429 0.77 1.9 12.3 0.35 0.457
CARB #313, 0.67 3.7 12.6 0.45 0.433 0.72 1.6 12.7 0.32 0.466
Average 0.64 4.3 12.3 0.46 0.431 0.76 1.8 12.5 0.34 0.463
c.o.v. 4% 13% 3% 4% 0% 4 .b 6% 2% 5% 1%
B20 #1/3 0.66 5.3 12.6 0.48 0.430 0.78 2.2 13.4 0.36 0.462
B20 #2/3 0.63 4.2 13.0 0.55 0.431 0.73 2.0 13.4 0.38 0.468
B20 #3!3 0.64 3.9 13.6 0.46 0.434 0:82 1.9 13.8 0.38 0.470
Average 0.64 4.5 13.1 0.50 0.432 0,78 2.0 13.5 0,37 0.467
c.o.v. 2% 17% 3% 10% 0% 6% 9% 20:6 3% 1%
C20 #1/3 0.63 4.2 12.9 0.49 0.434 0.75 1.9 13.1 0.36 0.468
C20 #213 0.63 3.9 12.8 0.48 0.431 0.80 1.8 13.3 0.39 0.473
C20 #313 0.67 3.9 12.8 0.48 0.432 0.73 1.8 13.1 0.37 0.467
Average 0.64 4.0 12.8 0.48 0.432 0.73 1.8 13.1 0.37 0.467
c.o.v. 4% 4% 1% 2% 0% 11% 3% 2% 5% 1%
Average Percent Changefirom Average EPA Cert. Diesel Baseline
CARB vs. EPA Cert. 1% -19% =1% 0 k -1% -7% -21% -2% -12% -1%
B20 vs EPA Cert. 1% -17% 6% 7% -1 % -5% -10% 5% -3% 0%
C20 vs EPA Cert 1% -26% 3% 4% 0% -11% -19% 2% -5% 0%
C20 vs. CARB 1% -8% 4% 4% 0% -4% 2% 4% 8% 1%
C20 vs. B20 0% -11% -2% -2% 0% -6% -10% -3% -2% 0%
[o074] These data indicate that B20 diesel blends result in increased NO,,
emissions
in both line-haul and switch duty cycles and increased exhaust opacity
(represented by
21
CA 02607478 2007-10-23
PM production) in line-haul duty cycle conditions. (With respect to
particulate matter,
the 2004 study concluded that the type of fuel utilized had little impact on
the amount
of particulates emitted because such emissions in two-cycle diesel engines are
domina-
ted by lubricating oil-generated components.) The foregoing testing appears to
show
that this is not necessarily true with respect to the present fuel
composition.
[0075] Each of the emissions characteristics of the present fuel composition
is highly
desirable, both individually and in combination. This is particularly true in
view of the
fact that railroad engine emissions are coming under scrutiny from
environmental
agencies such as the EPA. Because two-cycle diesel engines constitute the vast
majority
of engines in use on railroads throughout North America and because off-road
#2
petrodiesel generally is considered a relatively dirty fuel (i.e., its
combustion results in
large amounts of particulates, SO2r NO,, species, etc.), the availability of
alternative fuels
that can assist these engines in meeting more stringent emission standards is
highly
desirable.
[0076] While the present fuel composition has been found to provide
significant
emission benefits when used in two-cycle diesel engines, similar levels of
improvement
have not yet been proven in the four-cycle diesel engines more commonly
employed in
automobiles such as long-haul trucks; more accurately, initial, cursory
studies do not
appear to indicate that the improvements are as dramatic as those seen in two-
cycle
diesel engines. This could be due to any one or more of a variety of factors
including
pressure and temperature differences in the respective combustion chambers,
more
efficient cooling of combustion chambers in four-cycle engines, and the like.
[0077] However, the aforementioned cursory studies have shown other advantages
in the use of a fuel composition according to the present invention. For
example, tests
on a four-cycle CumminsTM turbo, inter-cooled diesel engine (set up to run at
2200 rpm,
which was believed to fairly approximate a freight truck at a constant
velocity of 24.6
m/s, i.e., 55 mph) using a commercially available #2 petrodiesel (A) and three
alterna-
22
CA 02607478 2007-10-23
tive fuels - a commercially available B20 biofuel (B), i.e., an 80:20 blend of
petrodiesel
and a methyl ester derivative of soybean oil, filtered waste vegetable oil
(C), and a fuel
composition according to the present invention (D) - resulted in the following
engine
efficiency data:
Table 5: Engine Efficiency, Four-Cycle Diesel Engine
A B C D
Engine efficiency* (%) 100 98.6 90.3 99.1
Coolant temp. increase ( C) 25 26 28 23
Fuel efficiency (km/L) 10.6 10.3 8.5 9.3
* Relative to petrodiesel.
The engine efficiency on each fuel was calculated using the formula xle =
1/(BFSCxLHV)
where rle is the engine efficiency for a given time interval, BFSC is the
brake specific fuel
consumption for that time interval, and LHV is the lower heating value of the
fuel. The
coolant temperature increase was determined after the engine had been warmed
as
close as possible to 71 C (160 F) before running it for 15 minutes with each
fuel type.
[0078] Based on the data of Table 3, a fuel composition of the present
invention
appears to result in a calculated engine efficiency that is far better than
that of waste
vegetable oil and even better than that of a commercial B20 diesel blend.
Also, use of a
fuel composition according to the present invention resulted in a coolant
temperature
increase less than that of all other fuels tested, including two commercial
fuels. With
respect to fuel efficiency, a fuel composition of the present invention
exhibited 10%
better results than a fuel made from waste vegetable oil; additionally,
assuming a linear
extrapolation of efficiency decrease with an increase in the percentage of
biofuel in the
blended product, the same fuel composition appears to yield -5 % better
results than a
biofuel presently considered commercially acceptable, i.e., a methyl ester
derivative of
soybean oil. This is contrary to reports of better power and consumption
results for
23
CA 02607478 2007-10-23
methyl esters relative to equivalent ethyl esters; see, e.g., the Peterson et
al. article and
publications cited therein (all of which compare "pure" biofuels, i.e.,
biofuels not
containing significant amounts of free alcohol).
[0079] At least the last of the foregoing efficiency results is somewhat
surprising in
view of the fact that the present fuel composition includes a not significant
amount of
an alcohol such as ethanol. The presence of such alcohols in a petroleum-based
fuel
typically would be expected to yield reduced fuel efficiency values. However,
the fore-
going results seem to indicate that the present fuel composition provides
better fuel
efficiency than "pure" biodiesel products.
[0080] Use of the present fuel composition does not appear to require the use
of any
special equipment or to require the modification of existing engine equipment.
Specifically, some literature and manufacturer warranty information appears to
indicate
that special seals and gaskets are required if a neat biofuel is to be run in
an engine.
However, none of the testing to date has shown this to be necessary with the
fuel
composition of the present invention.
[0081] Conversely, use of the present fuel composition has been shown to have
at
least one positive effect on equipment in which it is used. Specifically,
combustion of
the composition fuel appears to provide a detergency effect to metal engine
parts.
Contrary to other studies (see, e.g., the Peterson et al. article) which have
seen either no
improvement in coking or even somewhat worse coking, use of the present fuel
compo-
sition in a two-cycle railroad diesel engine has resulted in metal parts
(e.g., fuel injectors
and cylinders) that are far cleaner than before they were prior to use of the
present fuel
composition.
[0082] At present, this effect is not fully understood. Specifically, whether
the fuel
composition actually provides a detergency effect that helps to remove
previous
deposits or whether the fuel composition merely reduces the amount of new
deposits to
a level that permits normal operation of the engine to remove prior deposits
(e.g.,
24
CA 02607478 2007-10-23
through vibrations) is not known. However, what can be said with some
certainty is
that engine parts that are coated with fewer deposits are expected to be
easier to cool
and to run more efficiently.
[0083] Terms and phrases used in this description are believed to assist in
the
understanding of the composition and processes of the present invention.
However, no
unnecessary limitations are to be implied from the brief, concise description
of
illustrative embodiments provided.
