0960-8524(95)00033-X
ELSEVIER
the triglycerides. This process leaves alcohol fattyacid esters, which have a viscosity far less than that
of oils and fats.
The esters of oils and fats can be directly blended
with diesel fuel. The advantage of blending esters
with diesel fuel is reduced emissions. As the EPA
imposes limits on exhaust emissions, these esters
should become increasingly attractive as a cleanerburning fuel. Research on blending alcohol esters of
different types of oils and fats in different ratios with
diesel fuel have been reported. Schumacher et al.
(1993) reported a reduction in carbon monoxide
(CO), hydrocarbons (HC) and smoke with an
increase in soydiesel (methyl esters of soybean oil)
concentration in the blend, whereas oxides of nitrogen (NOx) increased. Schlautman et al. (1986)
conducted a 200 h screening test using a 3:1 (v/v)
blend of unrefined, mechanically expelled, soybean
oil and No. 2 diesel fuel in a direct injection engine.
They had to terminate the screening test after 159 h
because the engine could not hold a constant load
and there was a 670% increase in the viscosity of the
lubricating oil. They further observed abnormal carbon deposits on all combustion chamber parts,
including the injectors. Schlick et al. (1988) evaluated the performance of a direct injection engine
with 1:3 (v/v) blends of soybean oil and sunflower
oil with No. 2 diesel fuel. They reported satisfactory
engine performance as far as power output, thermal
efficiency and lubricating oil data from the Engine
Manufacturer Association (EMA, 1982) screening
test was concerned, but when the general condition
of the combustion chamber and the fuel injectors
was investigated, heavy carbon deposits were discovered. Foseen et al. (1993) used methyl soyate
(from 0-40%) and diesel fuel blends in a transient
mode test of a DDC 6V-92 TA engine and found
that the addition of up to 40% methyl soyate did not
affect peak torque, but there was a small drop in
power at the 40% level of substitution. They reported a reduction in CO, HC and particulate matter
Abstract
238
and an increase in NOx emissions. They recommended use of 20% methyl soyate blend with diesel fuel.
The purpose of this investigation was to optimize
the blend of No. 2 diesel fuel, methyl tallowate and
ethanol to reduce emissions without significantly
affecting engine performance.
METHODS
Engine and instrumentation
A Cummins N14-410 diesel engine was used in this
study. Specifications of the engine are presented in
Table 1.
The engine was coupled to an Eaton 522 kW (700
hp) dynamatic, eddy-current, dry gap dynamometer
(EATON Power Transmission Systems, Eaton
Corp., Kenosha, WI) with a DANA 1810 coupler.
Engine torque was measured with a load cell and a
Daytronic system 10 integrator (Daytronic Corp.,
Miamisburg, OH) and speed was measured using a
60-tooth sprocket and magnetic pick-up attached to
the dynamometer. Fuel consumption was measured
with an EM Corp. (Lincoln, NE) custom-built massmeasurement system, in which fuel weight was
measured over an operator-selected time period. Air
flow into the engine was measured with a Badger
BVT-IF venturi flow meter (Badger Meter, Inc.,
Tulsa, OK). The air flow meter was positioned in a
0"15 m diameter, 5.2 m long pipe with a surge tank
between the meter and the engine. A throttle valve
was used at the inlet of the surge tank to control
engine inlet pressure. The exhaust system consisted
of a 2-1 m length of 0.13 m diameter exhaust tubing
leading into a 0.25 m diameter duct to the outside
air. A centrifugal fan provided exhaust ventilation. A
throttle valve was positioned in the exhaust tubing to
control exhaust back pressure.
Temperatures of the exhaust of each cylinder, fuel
and coolant going into and out of the engine, and
crank-case oil, were measured using thermocouples
and a Daytronic System 10 coupled with an AutoNet
data-acquisition system. Pressures were measured
with analog gauges and manometers (H20 and Hg)
calibrated with a dead-weight tester.
Exhaust emission analyses were performed using
different analyzers for each of the exhaust gases.
Specifications
Type of engine
Horsepower (Rated)
Bore x stroke
Displacement
Compression ratio
Valves per cylinder
Aspiration
Turbocharger
239
Engine speed,
rpm
Load,%
Engine speed,
rpm
Load,%
1100
1200
1400
1600
1800
1900
100
100
100
100
100
100
1800
1800
1800
1800
1200
1200
1200
Idle
100
75
50
10
100
75
50
0
Fig. 1. Effects of engine speed and fuel blends on corrected power output.
Testing procedure
The engine was warmed-up at low idle long enough
to establish correct oil pressure and was checked for
any fuel, oil, water and air leaks. The speed was
then increased to 1600 rpm and a sufficient load was
applied to raise the coolant temperature to 71C.
After completion of a standard warm-up procedure,
the intake and exhaust restrictions were set at rated
engine speed (1800 rpm) and fall power and from
then on were not adjusted for different speeds or
loads after initial settings were completed.
The engine was run at the specific speeds and
loads for a minimum of 6 min and data were recorded during the last 2 min of operation. The response
variables included power, torque, brake specific fuel
consumption (BSFC), BSHC, BSCO, BSCO2,
BSNOx, BSO2 and brake specific smoke. These data
were recorded at 5 s intervals for 2 min and averaged over that period. After completion of one set
of experiments with four fuels the whole set was
replicated.
Optimization of fuel blends
Engine performance, corrected to SAE conditions,
and emissions analyses were conducted for each fuel
blend as described above. Statistical analyses for the
response of the engine with different fuel blends
were performed to determine the trends of the
response variables. The response variables considered were engine power output, torque, BSFC,
BSCO, BSCO2, BSHC, BSNOx, BSO2 and smoke.
The optimization was based on maximizing power
output and minimizing engine emissions. Response
surfaces for power, torque and BSFC and response
curves for emission characteristics using standard,
eight-mode, steady-state tests were plotted.
RESULTS AND DISCUSSION
A blend of ethanol and methyl tallowate was optimized to reduce the viscosity of methyl tallowate by
Ali and Hanna (1994a). They recommended a blend
of 65:35 methyl tallowate and ethanol, respectively,
to have a viscosity similar to No. 2 diesel fuel at
40C. The same blend of methyl tallowate and ethanol was used in this study. The viscosities of
80:13: 7, 70:19.5:10.5 and 60: 26:14 diesel: methyl
tallowate:ethanol blends were found to be 1"98, 1.97
and 2.01 mPa-s, respectively, at 40C as compared to
2.07 mPa-s for No. 2 diesel fuel at the same temperature. The calculated cetane index of methyl
tallowate was found to be 57"78, which reduced to
around 50 when blended with ethanol and diesel
fuel in different ratios. The calculated cetane index
of No. 2 diesel fuel was also found to be 50. Energy
content per unit mass of the diesel fuel was 45.51 kJ/
g, whereas that of methyl tallowate: ethanol (65 : 35)
blend was 36"16 kJ/g. The energy content of the
blends of diesel: methyl taUowate: ethanol reduced
proportionately as the percentage of methyl tallowate and ethanol increased in the blend.
Engine performance
The engine power outputs corrected to the SAE
standard J1349 (1992) at full load for all four test
fuels and six speeds are shown in Fig. 1. Statistical
analyses performed to find the effects of engine
speeds and fuel blends on power output showed that
the fuel blends had a significant linear effect
( F = 20"68, P r > F = 0.0001), whereas engine speed
had a significant fourth order polynomial effect
(F = 15.21, Pr>F = 0.0004). No interaction between
engine speed and fuel blend was observed. The
regression model for the power output, in the range
of 1100-1900, was
P = - 6505.96 + 18.66S- 0.0196S 2 + 0.91 x 10-6S3
-1"617 x 10-9S4 + 0"3365D
(R 2 = 0"94)
240
x 1 0 - 5 S 3 -8"3282 x 1 0 - 9 S 4 +
2"1414D
(R 2 = 0"9702)
where T = torque (Nm); S = engine speed (rpm);
and D = diesel content in fuel blend (%).
Maximum torque was produced at an engine
speed of 1200 rpm. At this speed there was a linear
drop in torque with an increase in the methyl talloware:ethanol content in the fuel blend. As in the
case of power output, the torque produced by the
engine also reduced by 1.03% each time, with a 10%
increase in the methyl tallowate:ethanol blend in
the fuel. The maximum torque of 2085 Nm was produced at 1200 rpm with No. 2 diesel fuel, which
dropped by 21.4 Nm each time another 10% of diesel was replaced with the methyl taUowate and
ethanol blend.
The BSFCs at full load for all fuel blends and
speeds are shown in Fig. 3. Statistical analyses
0o5
10
)8
COa
241
1
O.8
0
O.g
"i
o
o
CO
0.75
o
0.7
I
I
I
I
50:32.5:1"L5 (10:28:14 70:19.5:100 80:15:7
00:60:3.5 100:0:0
DlelOI Fuel : Methyl Tallowato : Ethanol Blend
S0:320:17.5
0.7
14
| .......
-e .......
O~
.e. . . . . . . . . . . . . . . . .
12
10
.!
OO
"9.
o"
"I" O.4
4
0.3
0.2
2
t
79:10.5:10.5 80:13:7
110:8.11:3.5 1000.'0
00".,28:14
Variations in BSCO and BSCO2 emissions for different fuel blends and eight-mode tests are shown in
Fig. 4. Regression analyses performed for the effect
of fuel blends on BSCO emissions showed that there
was a significant quadratic trend (F -- 9.86,
Pr>F=0.0184). The regression model for BSCO
emissions as a function of fuel blend was
BSCO = 1.0724-0.011144D + 9.3 x 10-5D2
(R 2 = 0.7977)
where BSCO = brake specific CO emissions (g/kWh) and D = diesel content in the fuel blend (%).
It was observed that BSCO emissions decreased
with an increase in methyl tallowate:ethanol blend
in the fuel. Maximum BSCO emissions of 0.8875 g/
kW-h were observed with No. 2 diesel, which was
well below the upper limit of 11.4 g/kW-h set by the
EPA (Brezonick, 1994).
The BSCO2emissions did not have any significant
trend with the fuel blends used in this study. Statistical analyses performed for BSCO2 emissions
showed that the slope of the regression line was
almost zero. It was concluded that BSCO2 emissions
do not depend on the fuel blend. In such a case the
242
6. A regression analysis performed on BSNOx emissions data did not show a statistically significant
trend. The mean value of the BSNOx emissions was
used to interpret the results. The mean BSNOx
emission was 6.33 g/kW-h, as compared to the allowable 9.2 g/kW-h set by the EPA (Brezonick, 1994)
for a diesel engine of 130 kW size or more.
Bosch smoke units are an indication of particulate
and soot formation in the exhaust of an engine.
Particulates contain primarily carbon particles and
some unburned HCs. The observed smoke readings,
in Bosch smoke units, were converted into soot concentrations (mg/m3) at 15C and 760 mm Hg using
conversion chart SAE J255a (SAE, 1994) and then
converted to soot and particulates (g/kW-h) for the
eight-mode test. The trend of smoke emissions with
fuel blends is shown in Fig. 6. A regression analysis
performed on smoke emissions data showed a significant linear effect (F = 453.06, P r > F = 0.0001) of
fuel blend. The regression model describing the
trend of smoke with fuel blend was
BSS = 0.605- 0.00365D
(R 2 = 0.9971)
CONCLUSIONS
1. Engine performance with a methyl tallowate:
ethanol:diesel fuel blend was not affected to a
great extent from that of diesel-fueled engine
performance. There was a 1.1% power reduction and a 1.03% torque reduction for each
10% replacement of diesel fuel with methyl tallowate: ethanol blend.
2. Brake specific fuel consumption was increased
by 2"37% for each 10% increase in the methyl
tallowate: ethanol blend in the fuel.
3. There was a significant reduction in BSCO
emission with an increase in the methyl
tallowate:ethanol content in the fuel blend.
The BSCO emission was always less than the
limit set by the EPA. There was no change in
BSCO2 emissions.
4. The BSHC emissions had a significant quadratic trend with fuel blend. Minimum BSHC
emissions were observed with the 80:13:7 diesel: methyl tallowate: ethanol blend.
5. The BSO2 emissions did not change with an
increase in methyl tallowate:ethanol content in
the blend.
6. There was no change in BSNOx emissions with
an increasing methyl tallowate:ethanol content
in the blends. The BSNOx emissions remained
statistically the same for all the fuel blends
used in this study and were always less than the
EPA's limit of 9.2 g/kW-h.
7. Smoke emissions increased linearly with an
increase in the methyl tallowate: ethanol content of the blends.
8. A blend of 80:13 : 7 diesel: methyl tallowate:
ethanol should be used to minimize emissions.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the contributions of Kevin G. Johnson, Lab Technician,
REFERENCES
243