Science of the Total Environment 408 (2010) 1050–1058

Contents lists available at ScienceDirect

Science of the Total Environment

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v

Experimental study on particulate and NOx emissions of a diesel engine fueled with
ultra low sulfur diesel, RME-diesel blends and PME-diesel blends
Lei Zhu, Wugao Zhang, Wei Liu, Zhen Huang ⁎
School of Mechanical Engineering, Shanghai Jiaotong University, Shanghai, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Ultra low sulfur diesel and two different kinds of biodiesel fuels blended with baseline diesel fuel in 5% and
Received 10 June 2009 20% v/v were tested in a Cummins 4BTA direct injection diesel engine, with a turbocharger and an inter-
Received in revised form 15 October 2009 cooler. Experiments were conducted under five engine loads at two steady speeds (1500 rpm and 2500 rpm).
Accepted 20 October 2009
The study aims at investigating the engine performance, NOx emission, smoke opacity, PM composition, PM
Available online 13 November 2009
size distribution and comparing the impacts of low sulfur content of biodiesel with ULSD on the particulate
emission. The results indicate that, compared to base diesel fuel, the increase of biodiesel in blends could cause
Keywords:
Biodiesel
certain increase in both brake specific fuel consumption and brake thermal efficiency. Compared with baseline
Ultra low sulfur diesel diesel fuel, the biodiesel blends bring about more NOx emissions. With the proportion of biodiesel increase in
Emission blends, the smoke opacity decreases, while total particle number concentration increases. Meanwhile the ULSD
SOF gives lower NOx emissions, smoke opacity and total number concentration than those of baseline diesel fuel. In
Sulfate addition, the percentages of SOF and sulfate in particulates increase with biodiesel in blends, while the dry soot
Particle size distribution friction decreases obviously. Compared with baseline diesel fuel, the biodiesel blends increase the total
nucleation number concentration, while ULSD reduces the total nucleation number concentration effectively,
although they all have lower sulfur content. It means that, for ULSD, the lower sulfur content is the dominant
factor for suppressing nucleation particles formation, while for biodiesel blends, lower volatile, lower aromatic
content and higher oxygen content of biodiesel are key factors for improving the nucleation particles
formation. The results demonstrate that the higher NOx emission and total nucleation number concentration
are considered as the big obstacles of the application of biodiesel in diesel engine.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction
As we know, millions of vehicles on the road are considered as
major contributors to the worse environment and shorter supply of
petroleum. In order to face the increasingly stringent emission regu-
lations and consumption of petroleum reserves in the near future,
more and more technologies have been explored to reduce the fuel
consumption and engine emissions (Agarwal, 2007). The use of bio-
fuel, such as alcohol and biodiesel, could partly replace petroleum fuel
consumption and reduce toxic emission and more importantly restrain
the life cycle emission, such as CO2 emissions. Because of its repro-
ducibility, nontoxicity, and sulfur free, biodiesel is considered as a
potential alternative fuel of diesel fuel. Furthermore, due to its similar
properties to diesel fuel, there is no need to modify the engine when
the engine is fueled with its blends, while for B100, it is necessary
to change the fuel system (Ramadhas et al., 2004). Biodiesel, called
methyl/ethyl ester, is produced through a transesterification process
with methanol or ethanol and its feedstock, such as vegetable oil,
⁎ Corresponding author. Tel./fax: +86 21 34206540.
E-mail address: z-huang@mail.sjtu.edu.cn (Z. Huang).
animal fat and waste cooking oil, which is becoming a hot topic
(Graboski and McCormick, 1998).
A large amount of studies have been conducted on the biodiesel
properties and emissions from biodiesel. Knothe and Steidley (2005)
reported that the molecule structure, such as chain length and
unsaturation might have influence on viscosity of biodiesel. Tat et al.
(2000) and Boehman et al. (2004) also did some experiments on bulk
modulus, and stated that the bulk modulus of biodiesel had influence
on the timing of injection. In general, compared to diesel fuel, most of
the literature showed that biodiesel could significantly reduce CO, HC,
and smoke, while increasing NOx. According to a review of published
emission data for heavy-duty engines by EPA (Environmental Protec-
tion Agency of USA, 2002), from diesel to B20 (20% biodiesel by
volume), CO, HC, and PM decreased by 13%, 20% and 20% respectively,
while NOx emission increases by 4% on average. The same trend was
obtained in an extensive review (Lapuerta et al., 2008a). Thus, the
higher NOx emissions of biodiesel might be considered as an obstacle
of biodiesel application. However, there are also opposite trends of
NOx emission in the literature. The results of Rakopoulos et al. (2006)
studies showed a slight decrease in NOx emission.
Because some components of particulate are toxic, it is necessary
to investigate the composition of PM. Previous composition studies

0048-9697/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.scitotenv.2009.10.056
 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058 1051

focused mainly on SOF fraction. Bagley et al. (1998) studied the SOF
and dry soot of PM, and reported that the proportion of SOF in PM was
increased and dry soot was decreased by using biodiesel, compared
with diesel fuel. However, there is nearly no literature about the
sulfate fraction of PM, when the diesel engine fueled with biodiesel
blends. The sulfur content of fuel plays a key role in the formation
of nucleation mode particles. During the combustion process the
majority of the fuel sulfur is converted to SO2, and then be oxidized
into SO3 further. An exothermic reaction leads to formation of sulfu-
ric acid aerosols, together with the chemical bonded water, which is
emitted as a particulate matter (Opris et al., 1993).
Over the past years, diesel exhaust particle size and number dis-
tribution have attracted the attention of the public. As particles be-
come smaller, it is much easier for them to diffuse in the ambient
air, and absorb more hazardous substances due to larger specific sur-
faces, then easier to be inhaled into the lungs (Donaldson et al., 1998;
Ferion et al., 1992). Diesel exhaust particles are typically composed
of nucleation particles, Dp b 50 nm; accumulation particles, 50 nm b
Dp b 1000 nm. Many studies have been conducted on the particle
number concentration and size distribution, when the engine fueled
with biodiesel (Tsolakis, 2006; Lapuerta et al., 2008b; Di et al., 2009).
They also concluded that the mean diameter of particles for biodiesel
tended to decrease. Krahl et al. (2001) stated that most of the par-
ticles for RME (rapeseed methyl ester) had smaller size in the range
of 10 to 40 nm.
One known method for reducing the total particle number con-
centration from the diesel engine is to use lower sulfur fuel, because
sulfur content is one of the key factors of the particle formation, espe-
cially the nucleation mode particles. Ristovski et al. (2006) studied
the influence of sulfur content in fuel on the nucleation particles'
formation. The results showed that there are two separate mechan-
isms responsible for the formation of nuclei mode particles. The first
mechanism was caused by semi-volatile organic compounds and sul-
furic acid at light engine load. The second was governed by sulfuric
acid and water nucleation at high engine load. However, there are
limited published results about the particle size distribution and com-
position of different sources of biodiesel and different sulfur content
fuels. Also the biodiesel is a sulfur free fuel, whose sulfur content is
even less than 10 mg/kg, according to Table 2, which shows possible
positive effects on the reduction of PM. On this account, it is necessary
to study the impact of biodiesel on PM and NOx emissions of diesel
engines, and to compare the impacts of biodiesel and ULSD with low
sulfur contents on the particles emission. Thus, in this investigation,
we will focus on the comparison of engine performance, NOx emis-
sion, smoke opacity, PM composition, PM size distribution and nucle-
ation mode particles with different fuels.
1.1. Test engine and fuel properties
A Cummins 4BTA direct injection diesel engine, with a turbocharg-
er and an intercooler, was used in this study. The main specifications
of engine are shown in Table 1. An eddy-current dynamometer was
coupled with this engine, which was available for measuring and ad-
justing the speed and torque of the engine. A baseline diesel fuel, ultra
Table 1
Engine specifications.
low sulfur diesel, and two kinds of biodiesel were used in this study.
The major properties of these fuels are shown in Table 2, which were
tested by SGS Ltd. The baseline diesel fuel is a typical diesel fuel
similar to those available in China petrol stations. These two kinds of
biodiesel were produced from rapeseed oil and palm oil through the
transesterification with methanol, and here named as RME (rapeseed
methyl ester) and PME (palm methyl ester). The two blended fuels
were prepared by the volume ratio of 5% and 20% respectively, with the
baseline diesel fuel. Thus, there are six kinds of fuels tested in this
study, which are identified as diesel, ULSD, RME5, RME20, PME5, and
PME20.
1.2. Experimental equipments and procedures
The experimental layout is shown in Fig. 1. The gaseous species
of engine exhaust were measured by an analyzer (AVL Digas 4000).
Smoke opacity was measured by a smoke meter (AVL 439), which was
represented by extinction coefficient k, which increases with the in-
crease of smoke concentration. The particle size distributions were
measured with a TSI 3034 SMPS (scanning mobility particle sizer, TSI
Inc., Shoreview, MN) within a range of 10–487 nm. The SMPS 3034 is
composed of a DMA (differential mobility analyzer) and a TSI 3010
CPC (condensation particle counter). In order to dilute and cool the
exhaust sample, the exhaust gas of the engine was diluted with the
ambient air by the single staged dilution tunnel firstly, before passing
through the SMPS. A valve was set up to adjust the fraction of the
exhaust passing through the dilution tunnel. In order to prevent de-
position of solid particles and condensation of volatile organics, a
temperature controller was used to maintain the surface temperature
at 300 °C. Also the PM sample was collected with a glass fiber filter for
offline analysis. The proportion of SOF (soluble organic friction) was
extracted by dichloromethane and analyzed with the microwave ex-
traction method, while dry soot and sulfate were measured by carbon
and sulfur analytical instrument. The dilution ratio (DR) was calcu-
lated based on the following equation:
½CO2 exhaust −½CO2 background
DR =
½CO2 diluted −½CO2 background
where [CO2]exhaust is the CO2 concentration in the exhaust before
dilution, [CO2]diluted is the CO2 concentration in the mixture gas after
dilution, and [CO2]background is the CO2 concentration in the background.
The tests were conducted at two steady engine speeds (1500 rpm
and 2500 rpm), and at engine loads of 0 Nm, 75 Nm, 150 Nm, 225 Nm,
and 300 Nm, corresponding to brake mean effective pressures of
0 MPa, 0.24 MPa, 0.48 MPa, 0.72 MPa, and 0.96 MPa, respectively. At
each engine operating condition, the experiments were operated with
diesel fuel, ULSD, RME5, RME20, PME5, and PME20, respectively.
Moreover, for minimizing the cross contamination of different fuels
on each other, before each test, the engine was allowed to operate
with new test fuel for 30 min to clean the fuel system. In order to
assure the repeatability and comparability of the measurements for
Table 2
Properties of baseline diesel fuel, ULSD, RME, and PME.
Property Diesel ULSD RME PME

Cetane number 51 72 53 64
Model Cummins 4BTA
Lower heating value(MJ/kg) 42.5 43 39.8 40.0
Type In-line 4-cylinder Density(kg/m3) at 20°C 834 823 873 878
Suction type Turbo intercooler Viscosity(mPa s) at 20°C 3.03 3 6.38 7.11
Maximum power 88 kW/2800 rpm Heat of evaporation (kJ/kg) 280 250–290 300 300
Maximum torque 380 Nm/1600 rpm Boiling point (°C) 180–330 175–345 300–350 300–350
Bore х stroke 102 mm × 120 mm Carbon content (% wt) 87.4 86.6 76.8 76.6
Displacement 3920/cc Oxygen content (% wt) 0 0 10.5 11.2
Compression ratio 17.5:1 Sulfur content (mg/kg) 400 b1 b 10 b 10
Injection pump type In-line Bosch P7100 Aromatic content (% wt) 27.7 0.6 0 0
1052 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058

Fig. 1. Schematic diagram of experimental system.

different operating modes, the cooling water temperature was auto-
matically controlled by a temperature system, maintained at 85 °C,
and held within ±2 °C, while the oil temperature was kept at 90–
95 °C. Each test was repeated twice to ensure that the results are
repeatable within the experimental uncertainties. The experimen-
tal uncertainty has been calculated by the method of Kline and
McClintock (1953) and Ames et al. (2000). Before data collection,
the engine was allowed to operate under certain test mode for a few
minutes until the data became steady, especially the lubricant tem-
perature and cooling water temperature. In addition, the diesel engine
in the experiment was not modified during all the tests. The engine
performance, NOx emission, smoke opacity, PM composition, PM size
distribution and nucleation mode particles were studied with diesel,
ULSD, RME blends and PME blends under five loads at two steady
speeds.
more fuel injected in diffusion stage, which leads to lower BTE. More-
over, due to the larger density and viscosity of biodiesel, the spray and
atomization could be deteriorated. Fang et al. (2009), Fang and Lee
(2009) found that the biodiesel had the longer liquid penetration
and stronger fuel impingement than the diesel fuel, due to the higher
boiling point and higher density of biodiesel. Thus, the combustion
process could be worsened, and result in lower BTE. Meanwhile, oxy-
gen contents in blends will improve the combustion process and re-
sult in an increase of BTE. Thereby, these adverse factors for biodiesel

2. Results and discussion

2.1. Engine performance

The brake specific fuel consumption (BSFC) and the brake thermal
efficiency (BTE) can be calculated by the engine torque, the engine
speed and the mass consumption rate of the fuel. The BSFC decreases
with an increase in the engine load for all different fuels. For each test
mode, the higher the proportion of biodiesel in blended fuels is, the
higher BSFC it has, due to the lower heat value of biodiesel. The ULSD
gives the lowest BSFC within all fuels. Compared with ULSD, the BSFC
decreases 7.5% on average at 1500 rpm, based on different engine
loads and different fuels, while at 2500 rpm, there is no significant
decrease. Fig. 2 shows the variation of brake thermal efficiency (BTE)
under different loads of 1500 rpm and 2500 rpm. As it is shown in the
figure, for each engine speed, BTE increases with the increase in load
from 0.02 MPa to 0.96 MPa, while the increase of BTE is slight from
0.72 MPa to 0.96 MPa. Compared with 1500 rpm, the BTE of 2500 rpm
is lower due to increased incomplete combustion and mechanical
frication losses. At each test mode, the BTE of different fuels is almost
at the same level as each other except ULSD, which displays a better
performance in brake thermal efficiency. The maximum BTEs of diesel,
ULSD, RME5, RME20, PME5, and PME20 are 35.4%, 36.0%, 35.2%, 35.0%,
34.3% and 34.8%, respectively, at 2500 rpm. Statistical treatment of
data indicates that the maximum BTE of each fuel is significantly
different at the 95% confidence level (P b 0.05) with each other. Many
factors contribute to the variation of BTE. When the engine is fueled
with blended fuels, the decrease in low heating value discussed in the
previous paragraph and the increase in cetane number will lead to Fig. 2. Effect of different fuels and engine load on brake thermal efficiency.
 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058
blends interact with each other, and lead to the same level of BTE,
compared with diesel fuel.
2.2. NOx emissions
Fig. 3 shows the variation of NOx emissions under different engine
loads of 1500 rpm and 2500 rpm for different fuels. The NOx emissions
increase with engine load for all fuels. Compared with 2500 rpm,
the NOx emissions are much higher at 1500 rpm. For ULSD, the NOx
emissions are the lowest among different fuels at any certain modes.
However, Tan et al. (2009) stated that the fuel sulfur content had little
effect on the NOx emissions. Moreover, with biodiesel proportion
increasing, the NOx emissions increase simultaneously. The higher the
load is, the larger the gap between the diesel and biodiesel is. From
diesel fuel to PME20 at 0.96 MPa of 2500 rpm, the NOx emission
increases by 8%. Statistical treatment of data indicates that NOx
emissions of RME-diesel and PME-diesel blends are significantly
(P b 0.05) higher than that of the diesel fuel. Marshall et al. (1995)
investigated the NOx emissions in a Cummins diesel engine fueled
with biodiesel blends. They found a 3.7% increase of NOx emission
with the 20% biodiesel blend. Canakci (2005) conducted the emission
test in a diesel engine fueled with blends of soy bean oil biodiesel and
diesel, and blends of yellow grease biodiesel and diesel. They con-
cluded that, compared with diesel fuel, the NOx emissions increase
13.1% and 11.6% for soybean oil biodiesel and yellow grease biodiesel,
respectively. Moreover, the PME blends give higher NOx emissions
than those of RME blends, due to the higher viscosity and oxygen
content of PME compared with RME, according to the Table 2.
It is well known that the major factors of the NOx formation are
high combustion temperature, rich local oxygen concentration, and
Fig. 3. Effect of different fuels and engine load on NOx emission.
1053
long residence in the high temperature atmosphere. Obviously, with
biodiesel, the combustion temperature in chamber could be increased
and amount of oxygen is also increased, which could improve com-
bustion and lead to the formation of higher NOx emissions. Because of
the higher bulk modulus of biodiesel (Boehman et al., 2004), the quick
rise of pressure in injector leads to the early operating of nozzle and
advance injection, compared with that of diesel fuel. Lapuerta et al.
(2008c) concluded that the injection advance is attributed to the
increase of the NOx emissions. However, the cetane number of bio-
diesel is higher than diesel fuel. Higher cetane number means much
easier ignite and shorter ignition delay, which leads to the decrease of
the fuel amount in premixed combustion mode. The adverse factors
act against each other and lead to the increase of NOx emissions. How-
ever, Canakci (2005) and Lapuerta et al. (2008c) thought the oxy-
gen content in biodiesel has no obvious influence in NOx emissions
increase.
2.3. Smoke opacity
Fig. 4 shows the variation of the smoke opacity from the engine
fueled with different fuels at two steady engine speeds. According to
the figures, the smoke opacity increases with increase in loads. This is
acceptable due to the fact that more fuel is supplied to compensate the
high load; the more fuel is combusted in the diffusion phase at high
load than that at low and medium load, leading to the increase of
smoke opacity. The smoke opacity of 2500 rpm is higher than that of
1500 rpm. According to Fig. 2, the lower BTE at 2500 rpm means that
the combustion deteriorates at high engine speed, leading to the
higher smoke opacity. The ULSD gives lower smoke opacity than that
of diesel fuel at all test modes. At 0.48 MPa of 2500 rpm, the smoke
Fig. 4. Effect of different fuels and engine load on smoke opacity.
1054 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058
opacity of ULSD decreases by 28%, in comparison with baseline diesel
fuel. This might be due to the lower sulfur content of ULSD than that of
diesel fuel. Tan et al. (2009) studied the emissions of diesel engine
with different sulfur content fuels. They reported a downward trend
with the fuel sulfur content decrease for the same engine operating
condition and stated that fuel sulfur has more effect on the smoke
emission. Also, according to Table 2, the aromatic content of ULSD is
extremely lower than that of baseline diesel fuel. When the engine
fueled with ULSD, the reduction of smoke was attributed to the re-
duction of the aromatic compounds, which are known as soot pre-
cursors (Corporan et al., 2005).
It also can be concluded that the smoke opacity decreases, when
the engine is fueled with RME and PME blends. With increasing bio-
diesel proportion, the smoke opacity decreases remarkably. A sig-
nificant decrease is found at higher loads. From diesel fuel to RME20
and PME20, the smoke opacity decreases by 28.26% and 28.80%, re-
spectively, on average based on different engine loads at 2500 rpm.
Moreover, the smoke opacity of RME20 and PME20 are even lower
than that of ULSD at light engine loads. Ozsezen et al. (2008) con-
ducted a test on a naturally aspirated diesel engine fueled with bio-
diesel blends. They obtained 0.89% reductions with 5% biodiesel
blends, while 2.9% reduction with 20% biodiesel blends. Lapuerta et al.
(2008b) stated that the presence of the bonded oxygen could explain
the decrease of smoke, which reduces the probability of soot nuclei
formation in locally rich zones. When sufficient oxygen is available,
soot precursors species react with oxygen or oxygen-containing radi-
cals (like OH, O) and eventually produce CO rather than aromatics and
soot (Mueller et al., 2003). Moreover, the increase of oxygen content
in blends means a decrease of carbon content, which is the major con-
tributor of soot formation. Corporan et al. (2005) carried out a study
on a diesel engine fueled with biodiesel. He reported that these re-
ductions of smoke were due to the dilution of aromatics, which were
considered as the soot precursors. At the meantime, the start of in-
jection (SOI) was advanced with use of biodiesel stated in previous
section, due to the higher density and bulk modulus of compressi-
bility. The advancing SOI timing for biodiesel improved the smoke
opacity (Tsolakis, 2006; Ozsezen et al., 2009). Although the lower
sulfur content of biodiesel could help to reduce the smoke opacity to
some extent, the reduction of smoke opacity is attributed mostly to
the reduction of aromatics and to the presence of oxygen in the ester
molecules. Moreover, because of the higher oxygen content of PME,
the PME blends give lower smoke opacity than that of RME blends.
2.4. Particulate composition
It is well known that the diesel particle consists of soluble organic
friction, dry soot and sulfate. In this study, PM samples were collected
with glass fiber filters for offline analysis to obtain the composition
information about the effect of biodiesel on the SOF (soluble organic
friction), sulfate, and dry soot friction. The total PM mass on the filter
was determined from the difference in mass before and after expo-
sure to the diluted exhaust. The proportion of SOF was extracted
by dichloromethane and analyzed with the microwave extraction
method, while dry soot and sulfate were measured by carbon and
sulfur analytical instrument. And it was necessary that the samples
were stored under – 18 °C to 4 °C, until they were analyzed. The same
method for SOF analysis was reported in Bagley's study (Bagley et al.,
1998). Fig. 5 shows the proportion of SOF, dry soot, and sulfate for
different fuels under the steady condition, at 0.96 MPa of 2500 rpm.
At 0.96 MPa, the dry soot proportion of diesel, ULSD, RME5, RME20,
PME5, and PME20 are 80.15%, 91.77%, 77%, 55.3%, 78.01%, and 57.16%,
respectively. ULSD has the highest dry soot proportion within the
fuels, due to its highest carbon content. Also the dry soot proportion
decreases with increase in the percentage of biodiesel. Because of
the addition of biodiesel, the oxygen content in blends increases, while
the carbon content decreases in content. Thus, it is acceptable that the
Fig. 5. Effect of different fuels on particulate composition at 0.96 MPa of 2500 rpm.
proportion of dry soot decreases in the soot emission. In this study,
the decrease of smoke opacity is in-line with the reduction of dry soot
proportion because the smoke opacity is considered as a good indi-
cator of dry soot emission.
For ULSD, the SOF proportion is almost at the same level as that of
diesel fuel. While the SOF proportion of RME20 and PME20 is 4 times
and 3 times larger than that of diesel fuel respectively. It can be ob-
served from the figure that the SOF content depends on the biodiesel
content in the blended fuel. With the increase of the biodiesel content,
the proportion of SOF increases and the dry soot friction decreases at
all loads. In general, the origins of SOF in PM are partly of lubrication
oil, unburned oil, and compounds during combustion in the chamber
(Lin et al., 2008). Knothe et al. (2006) and Bagley et al. (1998) reported
that the higher SOF fraction observed with the biodiesel or biodiesel
blends could be attributed to the low volatility of biodiesel, which was
caused by the higher boiling point, higher viscosity and higher density,
leading to the worse mixture of fuel and air in chamber and incomplete
combustion. In the end, the unburned oil and compounds of uncom-
pleted combustion increase with biodiesel proportion, and result in
the increase of proportion of SOF in turn.
The sulfate proportions of diesel fuel, ULSD, RME5, RME20, PME5
and PME20 are 12.81%, 0.01%, 13.44%, 16.88%, 12.55%, and 17.94%,
respectively. It is acceptable that the sulfur content of fuel dominates
the content of sulfate in soot emission. As observed from the figure,
the ULSD gives nearly zero sulfate proportion because of its extremely
lower sulfur contents. After the addition of biodiesel, it is interesting
that the proportion of sulfate fraction increases. However, the sulfur
content in the biodiesel blended fuel decreases with the increase of
biodiesel fraction. The more the biodiesel is in the blended fuels, the
lower the sulfur content will be. It is questionable that the pheno-
menon of the increase of sulfate proportion belongs with the decrease
of sulfur content. The most possible explanation is that the observed
increase of sulfate proportion is just a consequence of the sharper
reductions in smoke and a slighter decrease in sulfate mass of bio-
diesel blends.
2.5. Particle number size distribution and total number concentration
The harmful effect of particle of diesel engines depends on its
mass concentration, but more importantly on its number concen-
tration. As particles become smaller, it is much easier for them to
diffuse in the ambient air, and absorb more hazardous substances,
then easier to be inhaled into the lungs (Donaldson et al., 1998; Ferion
et al., 1992). In this study, the SMPS was adjusted to measure the
particle diameter ranging from 10 nm to 487 nm. In order to ensure
the repeatable of SMPS measurement, each data was recorded three
times at each test mode. The particle number size distribution and
 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058 1055

the total number concentrations were tested at each selected mode.

Figs. 6(a–c) and 7(a–c) show the characteristics of particle number
size distribution with different fuels for engine loads of 0.24 MPa,
0.48 MPa and 0.96 MPa of 1500 rpm and 2500 rpm. The size distribu-
tion curves of all different fuels show similar bimodal structure, which
is composed of a nucleation mode (Dp b 50 nm) and an accumulation
mode (Dp N 50 nm). It can be found from the figures that the peak of the
curve increases slightly with the increase of engine load and engine
speed. All fuels have higher number concentration at nucleation mode
than at accumulation mode. At each test mode, the biodiesel blends
results in higher nucleation mode particle number concentration than

Fig. 7. Effect of different fuels and engine load on the particle size distribution of
2500 rpm.

Fig. 6. Effect of different fuels and engine load on the particle size distribution of
1500 rpm.
diesel fuel. With biodiesel fraction increasing, the peak of the curve
increases. The PME blends give a higher peak value of the curve than
that of the RME blends while, the biodiesel blends give lower accumu-
lation mode particle number concentration than diesel fuel and ULSD,
especially at high engine load of 0.96 MPa. Jung et al. (2006) studied
the characteristics of SME biodiesel fueled diesel particle emission and
concluded that the reduction of accumulation mode particles might be
due to a few factors. Firstly, the oxygen content of biodiesel could
reduce the soot formation, leading to the reduction of the cores for
accumulation mode formation. Moreover, the soot produced by bio-
diesel has a nanostructure with more reactive sites for oxidation,
1056 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058

which makes it much easier and faster to further burn and oxidize,
compared to soot made by their baseline diesel fuel. However, Cor-
poran et al. (2005) showed that the biodiesel had no apparent impact
on the particle size distribution. When the engine is fueled with the
ULSD, the peak of the curve holds the lowest level at nucleation mode.
Moreover, the difference of diesel and ULSD increases with the load,
which indicates that the most significant influence on the particle
number concentration is from the sulfur content and this can be ob-
served at the higher engine loads. Ristovski et al. (2006) stated the
same trend. Additionally, because of lower aromatic content of ULSD,
the soot formation could be suppressed and the particle number con-
centration could decrease. A detailed discussion about nucleation
mode will be shown in the next section.
In order to analyze the particle number concentration in detail,
Table 3 shows the total number concentrations under different engine
loads at two steady engine speeds. For each test fuel, the total particle
number concentrations increase with the engine load and speed. With
engine loads increasing, more fuel is injected into the chamber, lead-
ing to the increase of particle formation. This is mainly due to the
longer diffusion combustion duration and the reduced oxidation rate
of the soot in the expansion stroke (Tsolakis, 2006). Lapuerta et al.
(2007) stated that such local conditions under high load, with lower
air/fuel ratio, higher amount of fuel mass, lower excess oxygen, higher
pressure and temperature levels, not only contribute to soot nucle-
ation, but also promote the growth of existing soot nuclei.
For all test modes, the ULSD gives the lowest total number con-
centration among different fuels. Ristovski et al. (2006) and Rusyniak
et al. (2001) attributed this result to the extremely lower sulfur con-
tent of ULSD. It also can be found from the figure that when the diesel
engine is fueled with biodiesel blends, the total number concentration
increases. The total particle number concentrations of 20% biodiesel
blends might be approximately 2 times larger than that of the baseline
diesel fuel under different engine conditions. Due to the higher BSFC
of biodiesel blends, there are more particles formed when fuelled with
biodiesel blends. Tsolakis (2006) conducted a test in a single cylinder
engine with rapeseed oil biodiesel and diesel fuel at three steady
modes; he attributed the increase of particle number to injection
advance and injection pressure increase. For this reason, it is also
acceptable that the PME blends give higher total number concentra-
tion than that of RME blends, due to the higher viscosity of PME. Mathis
et al. (2005) also stated that advancing the start of injection could
reduce the diameter of primary soot particle. Additionally, compared
with diesel fuel, the lower volatility of biodiesel could nucleate and
condense into particles while the exhaust gas sample is cooled down in
the tailed pipe, which could increase the particles in nucleation mode
(Di et al., 2009). However, the higher oxygen content of biodiesel could
improve the combustion process in locally rich diffusion combustion
phase, which could lead to reduction of large diameter particles. Thus,
the increase of the total number concentrations of biodiesel blends is
dominated by nucleation mode particles.
2.6. Nucleation mode
It is widely known that the diesel engine particles are harmful to
human health and the environment, especially the nano-particles. As
particles become smaller, it is much easier for them to diffuse in the
ambient air, and in turn much easier to be breathed into the lung
(Donaldson et al., 1998). It is necessary to understand the formation
of particles in exhaust. The typical diesel exhaust particles contain
nucleation mode and accumulation mode. The nucleation mode is
made up of particles with a diameter below 50 nm of diesel exhaust
particles. During the dilution and cooling process, the nucleation of
sulfuric acid and low volatile organics occur. Meanwhile, the semi-
volatile organics could condensate on the surface of the existing nucle-
ation particles. The accumulation mode is made up of particles with
a diameter above 50 nm. These particles are composed of primarily
solid soot as cores and various semi-volatile and nonvolatile organics
as condensates (Bagley et al., 1998; Schneider et al., 2005). It is pos-
sible that the smaller the particles are, the more toxic they will be, due
to the higher specific surface area of small particles. In this study, the
total number of nucleation concentrations were measured and com-
pared for different fuels.
Fig. 8 shows the change of total nucleation number concentra-
tions with biodiesel blends and ULSD, compared with baseline diesel
fuel, at engine loads of 0.24 MPa and 0.96 MPa of 2500 rpm. The total
nucleation number concentration of ULSD decreases obviously, com-
pared with baseline diesel fuel. Ristovski et al. (2006) also tested par-
ticle emissions from twelve buses with low sulfur (500 ppm) and
ultralow sulfur (50 ppm) diesel fuel. They found that 74% of the par-
ticles were smaller than 50 nm with low sulfur diesel fuel, while 43%
with ultralow sulfur diesel fuel at the high engine load. Thus, for ULSD,
the lower sulfur content is the major contributor to the decrease of
total nucleation number concentration, compared to diesel fuel.
When the engine is fueled with biodiesel blends, the total nucle-
ation number concentration increases as the biodiesel percentages
increase, at each certain mode. For example, at the engine load of
0.96 MPa, the total nucleation number concentration of biodiesel
blends increases 12%, 57%, 25%, and 55%, respectively for RME5,
RME20, PME5 and PME20, in comparison with baseline diesel fuel.
Krahl et al. (2001) investigated the influence of RMB100 on particle
size distribution and observed that the number of particles for bio-
diesel in the 10–40 nm range increased while the number of particles
above 40 nm decreased, in comparison with diesel fuel. According
to the discussion in previous paragraph, the sulfur content is one of
the major contributors to nucleation mode particles. With low sulfur

Table 3
Total number concentration of different fuels.

Total number #/cm3 0 MPa 0.24 MPa 0.48 MPa 0.72 MPa 0.96 MPa

n = 1500 rpm
Diesel 1.53E+ 08 2.50E+ 08 5.63E+ 08 5.71E+ 08 6.40E+ 08
ULSD 3.11E+ 07 4.57E+ 07 5.12E+ 07 6.76E+ 07 7.97E+ 07
RME5 1.96E+ 08 2.67E+ 08 7.00E+ 08 8.02E+ 08 9.31E+ 08
RME20 3.14E+ 08 4.08E+ 08 8.25E+ 08 8.76E+ 08 1.03E+ 09
PME5 2.15E+ 08 3.72E+ 08 7.13E+ 08 8.54E+ 08 9.84E+ 08
PME20 3.28E+ 08 5.87E+ 08 9.49E+ 08 1.05E+ 09 1.11E+ 09
n = 2500 rpm
Diesel 7.17E+ 08 7.70E+ 08 7.93E+ 08 9.37E+ 08 1.42E+ 09
ULSD 1.33E+ 08 1.47E+ 08 1.55E+ 08 1.83E+ 08 2.67E+ 08
RME5 7.87E+ 08 8.78E+ 08 8.97E+ 08 1.07E+ 09 1.58E+ 09
RME20 9.54E+ 08 9.84E+ 08 1.05E+ 09 1.54E+ 09 2.17E+ 09
PME5 8.18E+ 08 9.26E+ 08 1.02E+ 09 1.12E+ 09 1.75E+ 09
PME20 1.04E+ 09 1.28E+ 09 1.45E+ 09 1.57E+ 09 2.19E+ 09
Fig. 8. Effect of different fuels on the change of total nucleation number concentration.
 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058
content fuel such as biodiesel blends and ULSD, suppressing the homo-
genous nucleation of sulfuric acid can reduce the nucleation mode
concentration (Kwon et al., 2003; Rusyniak et al., 2001). However,
when biodiesel blends are used in diesel engines, the total nucleation
number concentration increases, compared with diesel fuel. On the
one hand, the smoke emission of biodiesel blends could be reduced,
due to the high oxygen content, low aromatic compounds content
and injection advance of biodiesel, which means reduction of cores
for accumulation particles formation, and reduction of the number of
accumulation modes. The accumulation mode particles could absorb
the semi-volatile components and suppress the formation of nucle-
ation mode particles. If the accumulation mode particles are reduced,
the suppression of nucleation process turns weak, which enhances the
formation of nucleation mode particles (Li and Huang, 2009). On the
other hand, the unburned biodiesel is low volatile, due to its higher
boiling point, as shown in Table 2. Some of the unburned biodiesel
nucleates and condenses into minute particles while the gas sample is
cooled down, leading to an increase of the particles in the nucleation
mode (Di et al., 2009). The result of SOF in the previous section could
prove that higher SOF fraction of biodiesel blends is due to the higher
total nucleation number concentration of biodiesel blends. Mean-
while, sulfuric acid and water could produce a lot of nucleation mode
particles in the dilution and cooling process. The lower sulfur content
of biodiesel could suppress this kind of formation of nucleation mode
particles. These combined factors lead to an increase of total nucleation
number concentration for biodiesel blends. It means that the lower
sulfur content of biodiesel blends is not the dominant factor of the
formation of the nucleation particles, while the lower volatile, lower
aromatic content and higher oxygen content of biodiesel were the key
factors to improve the formation of nucleation particles.
3. Conclusion
The characteristics of engine performance, NOx emission, smoke
opacity, particulate composition, particle number and size distribu-
tion were investigated with different fuels. The following conclusions
can be drawn.
(1) The use of biodiesel blends leads to an increase in brake specific
fuel consumption, while almost same level of brake thermal
efficiency. However, the ULSD has an outstanding performance
in BSFC and BTE.
(2) Compared with baseline diesel fuel, the biodiesel blends give
higher NOx emissions. Within all test fuels, the ULSD gives the
lowest NOx emissions.
(3) The smoke opacity of biodiesel blends decreases remarkably,
in comparison with baseline diesel fuel. The ULSD has lower
smoke opacity than that of baseline diesel fuel.
(4) The dry soot proportions decrease with increase in the pro-
portion of biodiesel. And the SOF and sulfate proportions in-
crease with the percentage of biodiesel. The ULSD gives the
highest dry soot proportion and the lowest sulfate proportion.
(5) Compared with baseline diesel fuel, the total number concen-
trations of biodiesel blends increase. For ULSD, the total num-
ber concentration is lower than that of diesel fuel and biodiesel
blends. Moreover, the difference of total number concentration
between ULSD and baseline diesel fuel is dominated by the
nucleation mode range, indicating that the sulfur content plays
a key role in the formation of nucleation mode particles.
(6) The total nucleation number concentration increases with the
increase of biodiesel proportion, while for ULSD, the total nucle-
ation number concentration decreases sharply, compared with
baseline diesel fuel.
Thus, the future study will be focused on reducing NOx emission
and total nucleation number concentration of biodiesel.
1057
Acknowledgement
The authors would like to thank the National Natural Science Fund
of China (50676056), Key Project of Chinese Ministry of Education and
Shanghai Jiaotong University for financial support to this study.
References
Agarwal AK. Biofuels (alcohols and biodiesel) applications as fuels for internal combus-
tion engines. Prog Energy Combust Sci 2007;33:233–71.
Ames WF, Cain G, Steele WG. Mathematics for mechanical engineers. Boca Raton, Fla:
CRC Press; 2000.
Bagley ST, Gratz LD, Johnson JH, McDonald JF. Effects of an oxidation catalytic converter
and a biodiesel fuel on the chemical, mutagenic, and particle size characteristics of
emissions from a diesel engine. Environ Sci Technol 1998;9:1184–91.
Boehman AL, Morris D, Szybist J, Esen E. The impact of the bulk modulus of diesel fuels
on fuel injection timing. Energy Fuels 2004;18:1877–82.
Canakci M. Performance and emissions characteristics of biodiesel from soybean oil.
Proc Inst Mech Eng D J Automob Eng 2005;219:915–22.
Corporan E, Reich R, Monroig O, DeWitt MJ, Larson V, Aulich T, et al. Impacts of biodiesel
on pollutant emissions of a JP-8—fueled turbine engine. J Air Waste Manage Assoc
2005;55:940–9.
Di Yage, Cheung CS, Huang ZH. Experimental investigation on regulated and unregu-
lated emissions of a diesel engine fueled with ultra-low sulfur diesel fuel blended
with biodiesel from waste cooking oil. Sci Total Environ 2009;407:835–46.
Donaldson K, Li XY, MacNee W. Ultrafine (nanometer) particle mediated lung injury.
J Aerosol Sci 1998;29:553–60.
Environmental Protection Agency of USA (EPA). A comprehensive analysis of biodiesel
impacts on exhaust emissions, Draft tech. report 2002; EPA420-P-02-001.
Fang T, Lee CF. Bio-diesel effects on combustion processes in an HSDI diesel engine
using advanced injection strategies. Proc Comb Inst 2009a;32:2785–92.
Fang T, Lin YC, Foong TM, Lee CF. Biodiesel combustion in an optical HSDI diesel engine
under low load premixed combustion conditions. Fuel 2009b;88:2154–62.
Ferion J, Oberdorster G, Penney DP. Pulmonary retention of ultrafine and fine particles
in rats. Am J Respir Cell Mol Biol 1992;6:535–42.
Graboski MS, McCormick RL. Combustion of fat and vegetable oil derived fuels in diesel
engines. Prog Energy Combust Sci 1998;24:125–64.
Jung H, Kittelson DB, Zachariah MR. Characteristics of SME biodiesel-fueled diesel particle
emissions and the kinetics of oxidation. Environ Sci Technol 2006;40:4949–55.
Kline SJ, McClintock FA. Describing uncertainties in single sample experiments. Mech Eng
1953;75:3–8.
Krahl J, Baum K, Hackbarth U, Jeberien HE, Munack A, Schutt C, et al. Gaseous compounds,
ozone precursors, particle number and particle size distributions, and mutagenic
effects due to biodiesel. Trans ASAE 2001;44:179–91.
Knothe G, Steidley KR. Kinematic visconsity of biodiesel fuel components and related
compounds. Influence of compound structure and comparison to petrodiesel fuel
components. Fuel 2005;84:1059–65.
Knothe G, Sharp CA, Ryan TW. Exhaust emissions of biodiesel, petrodiesel, neat methyl
esters, and alkanes in a new technology engine. Energy Fuels 2006;20:403–8.
Kwon SB, Lee KW, Saito K, Shinozaki O, Seto T. Size-dependent volatility of diesel nano-
particles: chassis dynamometer experiments. Environ Sci Technol 2003;37:1794–802.
Lapuerta M, Martos FJ, Herreros JM. Effect of engine operating conditions on the size
of primary particles composing diesel soot agglomerates. J Aerosol Sci 2007;38:
455–66.
Lapuerta M, Armas O, Rodríguez-Fernández J. Effect of biodiesel fuels on diesel engine
emissions. Prog Energy Combust Sci 2008a;34:198–223.
Lapuerta M, Rodríguez-Fernández J, Agudelo JR. Diesel particulate emissions from used
cooking oil biodiesel. Bioresour Technol 2008b;99:731–40.
Lapuerta M, Armas O, Herreros José M. Emissions from a diesel-biodiesel blend in an
automotive diesel engine. Fuel 2008c;1:25–31.
Li XL, Huang Z. Emission reduction potential of using gas-to-liquid and dimethyl ether
fuels on a turbocharged diesel engine. Sci Total Environ 2009;407:2234–44.
Lin YC, Lee WJ, Chao HR, Wang SL, Tsou TC, Chang-Chien GP, et al. Approach for energy
saving and pollution reducing by fueling diesel engines with emulsified biosolu-
tion/biodiesel/diesel blends. Environ Sci Technol 2008;42:3849–55.
Marshall M, Schumacher LG, Howell S. Engine exhaust emissions evaluation of a
Cummins L10E when fueled with a biodiesel blend, vol. 952363. SAE Tec Pap Ser;
1995.
Mathis U, Mohr M, Kaegi R, Bertola A, Boulouchos K. Influence of diesel engine com-
bustion parameters on primary soot particle diameter. Environ Sci Technol 2005;39:
1887–92.
Mueller CJ, Picket LM, Siebers DL, Pitz WJ, Westbrook CK, Martin GC. Effects of
oxygenates on soot processes in DI diesel engines: experimental and numerical
simulations, vol. 2003-01-1791. SAE Tec Pap Ser; 2003.
Opris CN, Gratz LD, Bagley ST, Baumgard KJ, Leddy DG, Johnson JH. The effects of fuel
sulfur concentration on regulated and unregulated heavy-duty diesel emissions,
vol. 930730. SAE Tec Pap Ser; 1993.
Ozsezen AN, Canakci M, Sayin C. Effects of biodiesel from used frying palm oil on the
exhaust emissions of an indirect injection (IDI) diesel engine. Energy Fuels 2008;22:
2796–804.
Ozsezen AN, Canakci M, Turkcan A, Sayin C. Performance and combustion character-
istics of a DI diesel engine fueled with waste palm oil and canola oil methyl esters.
Fuel 2009;88:629–36.
1058 L. Zhu et al. / Science of the Total Environment 408 (2010) 1050–1058
Rakopoulos CD, Antonopoulos KA, Rakopoulos DC, Hountalas DT, Giakoumis EG.
Comparative performance and emissions study of a direct injection diesel engine
using blends of diesel fuel with vegetable oils or bio-diesels of various origins.
Energy Convers Manag 2006;47:3272–87.
Ramadhas AS, Jayaraj S, Muraleedharan C. Use of vegetable oils as I.C. engine fuels—a
review. Renew Energy 2004;29:727–42.
Ristovski ZD, Jayaratne ER, Lim M, Ayoko GA, Morawska L. Influence of diesel fuel sulfur
on nanoparticle emissions from city buses. Environ Sci Technol 2006;40:1314–20.
Rusyniak M, Abdelsayed V, Campbell J, El-Shall MS. Vapor phase homogeneous nucle-
ation of higher alkanes: dodecane, hexadecane, and octadecane. 1. Critical super-
saturation and nucleation rate measurements. J Phys Chem B 2001;105:11866–72.
Schneider J, Hock N, Weimer S, Borrmann S, Kirchner U, Vogt R, et al. Nucleation
particles in diesel exhaust: composition inferred from in situ mass spectrometric
analysis. Environ Sci Technol 2005;39:6153–61.
Tan PQ, Hu YZ, Lou DM. Regulated and unregulated emissions from a light duty diesel
engine with different sulfur content fuels. Fuel 2009;88:1078–85.
Tat ME, Van Gerpen J, Soylu S, Canakci M, Monyem A, Wormley S. The speed of sound
and isentropic bulk modulus of biodiesel at 21 C from atmospheric pressure to
35 MPa. J Am Oil Chem Soc 2000;77:285–9.
Tsolakis A. Effects on particle size distribution from the diesel engine operating on
RME-biodiesel with EGR. Energy Fuels 2006;20:1418–24.