Fuel 112 (2013) 224235

Contents lists available at SciVerse ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Effects of injection parameters and EGR on exhaust soot particle

number-size distribution for diesel and RME fuels in HSDI engines
Lukasz Labecki a, Andreas Lindner b, Wolfgang Winklmayr c, Renate Uitz d, Roger Cracknell d,
Lionel Ganippa a,
a
Centre for Advanced Powertrain and Fuels Research (CAPF), School of Engineering and Design, Brunel University, Uxbridge, London UB8 3PH, UK
b
TAC, Auhofstrae 11 B, 1130 Wien, Austria
c
tapcon & analysesysteme Gmbh, 5026 Salzburg, Austria
d
Shell Global Solutions, Shell Technology Centre Thornton, Pool Lane, Ince CH2 4NU, UK

h i g h l i g h t s

 RME emitted lower concentration of soot particles than diesel.

 Increasing the injection pressure reduced the particle size and its concentration.
 EGR caused smaller particles to agglomerate to form larger size particles.
 The presence of oxygen in the fuel reduced the exhaust particle number concentration.

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

Article history: The exhaust soot particles number size distributions obtained from the combustion of diesel and RME
Received 11 December 2011 fuels were investigated in a high speed direct injection (HSDI) diesel engine for different engine operating
Received in revised form 8 April 2013 conditions viz., fuel injection pressure, injection timing, exhaust gas recirculation (EGR) and engine load.
Accepted 1 May 2013
An Electrostatic Mobility Spectrometer (EMS) was used for characterising the exhaust soot particle num-
Available online 29 May 2013
ber size distribution. Increasing the fuel injection pressure reduced the particle size and its number con-
centration in the accumulation mode under low and high load conditions, but an opposite trend was
Keywords:
observed to the particle number concentration in the nucleation mode under higher load operation.
Diesel
Rapeseed methyl ester
The effect of fuel injection timings on the particle number concentration was not clear and consistent
Size distribution between diesel and RME fuels under low load operation. Under high load operation, the overall particle
number concentration for RME decreased but for diesel only the nucleation mode decreased, while the
accumulation mode remained unaltered when the fuel injection timing was retarded. The addition of
EGR caused the particles to agglomerate and form larger size particles, which were observed mostly in
the accumulation mode. Under most of the engine operating conditions RME emitted lower soot particle
concentration than diesel under both nucleation and accumulation modes. The presence of oxygen in the
fuel has the potential to lower the exhaust particle number concentration in diesel engines.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Diesel engines are widely used owing to their high thermal ef-
ciency and low fuel consumption. Despite these benets, particu-
lates and NOx emitted from diesel engines are of major concern
for the environment. Though the global air fuel mixtures are lean
in diesel engines, the local equivalence ratios can be greater than
2. These locally rich regions within the combustion chamber leads
to favourable conditions for the formation of soot particles, which
eventually results in higher exhaust tail pipe soot emissions [1].
The soot emissions from diesel engines consist of millions of carbo-
Corresponding author. Tel.: +44 895 266721.
E-mail address: lionel.ganippa@brunel.ac.uk (L. Ganippa).
naceous solid particles, which combine to form aggregates and ap-
pear as visible black smoke, if they are in high concentration.
Stringent emissions regulations are proposed in Europe and US
and around the globe to control and reduce the soot particulates.
Expensive after treatments systems are used to meet these emis-
sion regulations. Effective working and regeneration of these after
treatment systems results in a penalty of fuel cost. The number
concentration of soot particles emitted from diesel engine has also
been a major concern. Euro 5 and Euro 6 aim to reduce PM emis-
sion to 0.005 g/km for passenger cars but the future legislation
has also introduced a limit on the particle number emissions [2].
The generation of soot in engines can be minimised by enhanc-
ing the mixing time to have a more premixed type of combustion,
and also by operating the engine at conditions where the global

0016-2361/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2013.05.013
 L. Labecki et al. / Fuel 112 (2013) 224235
in-cylinder temperatures are low enough to inhibit the in-cylinder
soot formation. These conditions of improved mixing can be
achieved through strategies such as high levels of EGR, high swirl
and through advanced or retarded injection timing, which are nor-
mally referred to partially premixed or low temperature combus-
tion [36]. The benet of low soot emissions from these
strategies are mainly limited to low and medium load operation,
limited information is available in the literature about the exhaust
soot particle number size distribution under these operating condi-
tions for diesel and RME fuels. So this work is focused towards
exploring the exhaust soot particle number and size measurements
under the strategies that favour low temperature combustion.
For the purpose of characterising soot, devices based on electro-
static soot particles classication technique have been widely used.
These systems are usually capable of detecting particle mobility
diameters in the range from 10 nm to 1000 nm and their number
concentration. It has been shown that the general soot particle
number size distribution from diesel engines has a bimodal type
of distribution [7,8]. The nucleation modes are described as those
particles in a diameter range below 50 nm and normally they peak
at about 1020 nm. Besides soot, volatile organic fractions and par-
ticles that are formed from super-saturation of un-burnt hydrocar-
bons in the exhaust system are present in the nucleation mode,
and the nuclei mode contains 120% of the particles mass and
around 90% of the particles number [9]. The accumulation mode
ranges in size from about 50 nm to 500 nm, in this mode, carbona-
ceous particles from the combustion grew up through the conden-
sation of hydrocarbons or agglomeration with other particles.
Particulate measurement techniques require simultaneous sam-
pling and dilution of engine exhaust soot particles before they
are actually classied in an electro static classier. The dilution
of the soot particles are carried out to prevent coagulation, conden-
sation and also to reduce the concentration to the limit that can be
handled by the electrostatic classier and particle counting de-
vices. Thus dilution ratio is one of the important factors that inu-
ence the measured particle size distributions and it is also
inuenced by other factors such as temperature of diluting gas,
residence time, ageing and relative humidity, as all these factors
cause signicant artefacts to the measurement of soot particles.
In the already published literature, it has been shown that at low
dilution ratios (about 2030), high number concentrations of par-
ticles were observed in the nucleation mode [911]. The ways
through which the aerosol is diluted are of importance than the ac-
tual dilution ratio itself. It is mixing and cooling of soot with dilu-
tion gas in a controlled way that preserves the soot from further
nucleation or coagulation. The process of mixing and cooling of
soot particles had been done in different ways in different systems
[12,13]. The most commonly used system for simulating atmo-
spheric dilutions is the constant volume sampler (CVS). This meth-
od can be used as a full exhaust ow dilution tunnel (FEFD-tunnel),
or partial exhaust ow dilution tunnel (PEFD tunnel). Another sys-
tem that is used for dilution is the ejector diluter, which is based on
the venturi nozzle principle, the compact design of this system al-
lows couple of units to be operated in cascade to achieve different
dilution ratios. The rotating disc dilutor discussed in [14] is capable
of achieving a wide range of dilution ratios from 1:30 to 1:1000 by
varying the speed of rotation, and this is an ideal dilutor for parti-
cles below 1000 nm. Lyyrnen et al. [15] proposed a dilution sys-
tem where the dilution gas ows through a porous tube and
mixes with the sampled exhaust before it is fed to the classier
and it is referred as porous tube diluter. In this work a similar type
of dilution system was used to preserve the engine out aerosol be-
fore it was allowed to pass through a differential mobility analyser
and a faraday cup electrometer.
The strategies that leads to improved spray mixing process and
their effects on the exhaust soot particle number size and size are
225
very limited. So this work is focused towards exploring the exhaust
soot particle number and their size measurements for the strate-
gies such as varying the fuel injection timing, fuel injection pres-
sures and the exhaust gas re-circulation (EGR) under low load
and high load operation for both diesel and RME fuels.
2. Experimental apparatus and procedures
2.1. Experimental setup
All measurements were carried out in a four cylinder high speed
direct injection (HSDI) diesel engine. The schematic of the experi-
mental setup is presented in Fig. 1 and the specications of the en-
gine are provided in Table 1. The injectors are designed to have six
holes with a nozzle hole diameter of 0.154 mm, and the fuel was
pressurised using a common rail system. The engine control unit
(ECU) and its software allowed controlling the engine parameters
on real time basis. The exhaust gas analyser (Horiba, Mexa
7170DEGR) was used to determine the level of EGR and the level
of dilution in the primary dilution system.
A detailed schematic of the exhaust soot particle number size
analyser (tapcon & analysesysteme, Electrostatic Mobility Spec-
trometer EMS VIE-11) is presented in Fig. 2. The EMS consists of
a neutralizer, differential mobility analyser (DMA) coupled to a far-
aday cup electrometer (FCE). The neutraliser (Am-241), with an al-
pha activity of 60 MBq was used in the EMS and this radioactive
source is covered with a 2 lm thick layer of gold/palladium. The
DMA used in this work is used to detect particles of sizes ranging
from 5 nm to 650 nm.
The soot was sampled from the exhaust and diluted in two
stages (primary and secondary dilution). In the primary dilution
system, the sampled soot was rst allowed to mix with the inert
dilution gas (N2) in a perforated concentric tube arrangement.
Introduction of the inert gas at an early stage of sampling pre-
vented further changes to the engine produced soot in the exhaust
at high exhaust gas temperatures. The primary dilution ratio (PDR)
for the home built system was determined by measuring CO2
simultaneously in the exhaust tail pipe before and after the sam-
pled soot had been mixed and diluted with N2 in the concentric
tube arrangement, and the PDR is calculated as described in [16]
as follows:
CO2m  CO2a
PDR
CO2e  CO2a
where CO2m is the concentration of CO2 in the mixture (after the
mixing process), CO2e is the concentration of CO2 in the exhaust gas
(before the mixing process) and CO2a is the concentration of CO2 in
the ambient air.
While the secondary dilution system is based on a closed loop
mixing tube dilute conguration and it is shown as a dilution probe
in Fig. 2, this unit is coupled to an Electrostatic Mobility Spectrom-
eter (EMS). The N2 diluted exhaust gas after passing through the
FCE are allowed to pass through an absolute lter in the EMS sys-
tem. The ltered gas are heated to dilute the aerosol again in a
closed loop secondary dilution system before they are allowed to
pass through the charger, DMA and FCE. It is possible get up to
ten different dilution ratios by using a combination of six critical
orices (CO1, CO2, CO3, CO4, CO5 and CO6) as shown in the sche-
matic (Fig. 2). Always a constant amount of ow Qs (2.57 L/min)
was drawn into the neutraliser, classier and the FCE through
the operation of the pump P2 in conjunction with the main critical
orice (CO) placed ahead of the pump as shown in Fig. 2. Different
dilution ratios were achieved through the operation of one or sev-
eral critical orices at different combinations to vary the dilution
gas ow to the sampling probe and this alters the exhaust ow
226 L. Labecki et al. / Fuel 112 (2013) 224235

Fig. 1. The schematic diagram of the engine experimental setup.

Table 1
Engine specications.

Multi-cylinder turbo-charged DI diesel engine

Displacement (cm3) 1998.23
Cylinder number 4
Compression ratio 18.2: 1
Bore (mm) 86
Stroke (mm) 86
Cod-rod length (mm) 155

Qe into the system as the whole system operated with a well-de-

ned amount of Qs owing within a close loop conguration. The
SDR was calculated by the following relation:
Qs
SDR
Qe
where Qs is the sampled aerosol ow and Qe is the aerosol ow
adjusted by the critical orices. The total dilution ratio of the whole
system is the multiplication of PDR and SDR.
In the EMS system the high voltage unit of the central electrode
of DMA was swept within a very short time, a general description
of principle of working of this differential mobility spectrometer
can be found in [17]. From the measured number size distribution
the cumulative number concentration (CNC) of soot particles can
be calculated by the following equation:
n 
X 
dN
CNC  d ln Di
i1
d ln Di

where n is the consecutive particles diameter, dN/dlnD(i) is the
particle number concentration and dlnD(i) is the width of the size
interval. The CNC in the nucleation and accumulation modes are
calculated using the same formula but for different size ranges.
2.2. Experimental procedure
The main parameters that were investigated in this work are
injection timing, injection pressure and EGR. The EGR rate was
determined by measuring CO2 from the exhaust and the inlet man-
ifold of the engine and by taking the ratio as CO2in/CO2exh. Where
CO2in is the concentration of CO2 in the inlet manifold and CO2exh
is the concentration of CO2 in the exhaust manifold. In this study,
Fig. 2. The schematic diagram of: (a) the Electrostatic Mobility Spectrometer and
(b) the dilution probe (adopted from EMS user manual).
EGR was used without an intercooler, therefore the inlet tempera-
tures were higher at higher levels of EGR and this can have an ef-
fect on the soot emissions.
The variations in the soot particle number size distribution for
two different kinds of fuels viz., rapeseed methyl ester (RME) and
diesel at different engine operating conditions (injection pressure,
EGR and injection timing) at two different engine loads 42.7 Nm
(2.7 bar BMEP) and 80 Nm (5.0 bar BMEP) at 2000 rpm were sys-
tematically investigated in this work. The fuel injection pressure
 L. Labecki et al. / Fuel 112 (2013) 224235
was varied from 800 bar to 1200 bar while the injection timing
was varied from 9 to 3 deg bTDC, and for these conditions the le-
vel of EGR was also varied. For each fuel injection timing the ef-
fect of injection pressure and EGR have been varied
independently to study their effects on the soot particle number
distribution. Detailed test matrix about the engine operating con-
dition for both fuels is presented in Table 2. Based on preliminary
investigations, the PDR was set at 2.32 by varying the N2 ow in
the primary dilution system and the SDR was set at 2 by xing
the Qe ow to 1.25 L/min. The aerosol was heated to 300 C in
the secondary dilution system, for these conditions a good repeat-
ability in soot particle number size distribution was obtained at
this dilution condition. Heating the aerosol sample with the dilu-
tion gas to 300 C also helps in reducing the amount of volatiles
entering into the classier. Under all operating conditions, the
particles number size distribution was measured thrice and the
average from these three scans are reported in this work. The re-
sults are presented in two main sections according to low and
high engine load operation for different injection pressures, EGR
levels and injection timing.
3. Results and discussions
3.1. Particles number size distribution (low load)
The inuence of fuel injection pressure on the exhaust soot par-
ticle number size distributions for diesel and RME fuels are pre-
sented in Fig. 3 and this data was presented for an injection
timing of 6 deg bTDC. Signicant difference was observed in the
size range from 5 nm to 30 nm between diesel and RME fuels.
The smaller size particles (510 nm) emitted from the combustion
of RME fuel are almost an order of magnitude less than the number
concentration of particles emitted by diesel in the nucleation
mode. This difference could be attributed to the variations in the
chemical compositions and the sulphur content between these
fuels. The particle number size distribution in the nucleation mode
(050 nm) itself was not affected by different injection pressures
for both fuels. The variations that were observed for different injec-
tion pressures in this size range were well within the uctuation
limits, similar trends were observed when the fuel was injected
at 9 and 3 deg bTDC. For the retarded injection timing of 3 deg
bTDC the differences in the number concentration between diesel
and RME in the nucleation mode appeared to be larger compared
to 6 deg bTDC, but the injection pressure itself did not cause any
appreciable differences to the number size distribution for 6 and
9 deg bTDC.
Both fuels show almost the same particles size distribution in
the diameter range between 30 nm and 50 nm in the nucleation
mode, and the injection pressure had no effect on these size ranges.
On the other hand it can be seen that the injection pressure has an
effect on the number size distribution in the accumulation mode
for particles which are above 50 nm. The number concentration
of particles as well as the maximum detectable size range de-
creases with an increase in injection pressure. Similar observations
were made for both fuels at different injection timings. Increasing
the injection pressure causes better entrainment of hot air, which
favours evaporation and improved mixing which results in lower
exhaust soot emissions [18].
Fig. 4 shows the soot particle number size distributions for
diesel and RME when operated under different levels of EGR for
an injection timing of 6 deg bTDC. The EGR percentages were var-
ied up to 40% in the case of diesel, and up to 70% for the case of
RME, additionally the data for the effects of EGR on particle num-
ber size distributions were also acquired for 9 and 3 deg bTDC.
The presented results for the effect of EGR on the exhaust soot
227
particle number size distributions for the injection timing of
6 deg bTDC are similar to that of 9 and 3 deg bTDC injection
timings.
It can be seen that the nucleation mode was more pronounced
for diesel compared to RME. The exhaust soot particle number con-
centration in the nucleation mode (<50 nm) was normally less for
RME when compared to diesel, and the soot particle number con-
centration decreases with EGR for both fuels, similar observations
have been reported in [19]. Under high level of EGR operation (at
40%) for diesel, the nucleation mode was not present and the num-
ber concentration in the nucleation mode size range was the low-
est. The peak from the nucleation mode particles disappears with
higher levels of EGR, this leads to unimodal size distribution with
one peak in the accumulation mode. It can be seen that an increase
in EGR causes higher exhaust soot particle number concentration
in the accumulation mode and larger size range of detected parti-
cles. The addition of 70% EGR for the case of RME resulted in the
highest exhaust soot particles number concentration but the de-
tected maximum size range of soot particles for RME was less than
that of diesel at 40% EGR. The EGR technique is well known for the
reduction of NOx emissions. However, it has an adverse effect on
the soot formation as the engine operates on richer mixtures
where the availability of oxygen in the combustion process is lim-
ited. These are favourable conditions for high soot formation and
incomplete soot oxidation. Higher amount of soot emitted in a gi-
ven exhaust volume due to high EGR creates a more likely environ-
ment for coagulation, accumulation, condensation of volatile
fractions on the particles and surface growth; all of these factors
result in the detection of higher concentration of larger diameter
particles in the accumulation mode.
The injection timing effects under each engine operating con-
dition described in Table 2 were studied. The conditions pertain-
ing to the injection pressure of 1000 bar at 0% EGR have been
presented in Fig. 5 to show the effect of different injection timings
on the exhaust soot particles number size distributions for diesel
and RME fuels. Under all engine operating conditions, RME always
had lower exhaust soot particle number concentration in the
nucleation mode. The injection timing effect in this mode was
not very clear and consistent between RME and diesel. Retarding
the injection timing caused a slight increase in particle number
concentration for diesel in the nucleation mode, but an opposite
trend was observed for RME in this mode. In most cases the dif-
ferences in number concentration were within the uctuation
limits. In the accumulation mode, RME showed a reduction in
particle number concentration and a slight decrease in the maxi-
mum detection diameter range for retarded injection timing un-
der all engine operating conditions. But the retarded injection
timing did not show any signicant effect on the exhaust soot
particle number concentration and the detected particle diameter
range in the accumulation mode for diesel under low load opera-
tion. Generally, the retarded injection timing causes a shift to the
combustion phase towards the expansion stroke. This conse-
quently decreases the in-cylinder pressure and eventually the
in-cylinder temperature. In some cases, by retarding the injection
timing further results in a simultaneous reduction of soot and
NOx emissions, due to low temperature combustion [4]. The pres-
ence of oxygen in RME reduced soot formation under low temper-
ature operation and this leads to a reduction in particle number
concentration in the accumulation mode. Under low load opera-
tion, the soot particle diameters were usually in the range from
5 to 110 nm for all fuels and engine operating conditions. The
analysis from the inuence of these three engine operating
parameters under the low load operation have shown that EGR
has the strongest effect on soot particles size distributions by en-
tirely changing the soot particle size and number concentration
for both fuels.
228 L. Labecki et al. / Fuel 112 (2013) 224235

Table 2
Engine operating conditions.

42.7 Nm @ 2000 rpm DIESEL RME

effect of injection pressure
bTDC inj. press @ 0% EGR
9 deg
800, 1000, 1200
effect of EGR
effect of injection timing

EGR @ 800 bar

0%, 15%, 30% 0%, 15%, 30%, 40%, 50%, 70%
effect of injection pressure
inj. press @ 0% EGR
bTDC
6 deg

800, 1000, 1200

effect of EGR
EGR @ 800 bar
0%, 15%, 30% 0%, 15%, 30%, 40%, 50%, 70%
effect of injection pressure
inj. press @ 0% EGR
bTDC
3 deg

800, 1000, 1200

effect of EGR
EGR @ 800 bar
0%, 15%, 30% 0%, 15%, 30%, 40%, 50%, 63%
80 Nm @ 2000 rpm DIESEL RME
effect of injection pressure
inj. press @ 0% EGR
bTDC
9 deg

800, 1000, 1200

effect of EGR
effect of injection timing

EGR @ 800 bar

0%, 15%, 30% 0%, 15%, 30%, 33%
effect of injection pressure
inj. press @ 0% EGR
bTDC
6 deg

800, 1000, 1200

effect of EGR
EGR @ 800 bar
0%, 15%, 25% 0%, 15%, 30%, 33%
effect of injection pressure
inj. press @ 0% EGR
bTDC
3 deg

800, 1000, 1200

effect of EGR
EGR @ 800 bar
0%, 15%, 25% 0%, 15%, 30%, 32%

Fig. 3. Inuence of injection pressure on exhaust soot particles number size Fig. 5. Inuence of fuel injection timing on exhaust soot particle number size
distributions for diesel and RME at low load engine operation. distributions for diesel and RME at low load engine operation.

3.2. Cumulative particle number concentration (low load)

Soot particles of size smaller than 50 nm have been accounted

Fig. 4. Inuence of EGR on exhaust soot particle number size distributions for diesel
and RME at low load engine operation.
for nucleation mode and the particles of size larger than 50 nm
were accounted as accumulation mode. Fig. 6a6d presents the ef-
fects of injection pressure, injection timing and EGR on the cumu-
lative number concentrations in the nucleation and accumulation
modes for both fuels.
The injection timing and injection pressure did not have any
signicant inuence on the cumulative number concentration of
particles in the nucleation mode for diesel and RME fuel (Fig. 6a).
When compared to diesel, a reduction in the cumulative particle
number concentration was observed for RME in the nucleation
mode. The cumulative particle number concentration in the accu-
mulation mode seems to decrease slightly with higher injection
pressure, and this effect was more clearly seen for the retarded
injection timings and for higher fuel injection pressures. The
cumulative number concentration of particles in the accumulation
 L. Labecki et al. / Fuel 112 (2013) 224235 229

800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200
inj. pressure, bar inj. pressure, bar inj. pressure, bar

Fig. 6a. Cumulative number concentration of particles emitted at low engine load operation for diesel and RME fuels in the nucleation mode at different injection pressure
and injection timing.

800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200
inj. pressure, bar inj. pressure, bar inj. pressure, bar

Fig. 6b. Cumulative number concentration of particles emitted at low engine load operation for diesel and RME fuels in the accumulation mode at different injection pressure
and at different injection timing.

mode is signicantly lower when compared to the nucleation
mode for both fuels, about two orders of magnitude difference
can be seen in Figs. 6a and 6b respectively.
Fig. 6c shows the effect of EGR on the cumulative number concen-
tration of particles in the nucleation mode for both fuels. As the per-
centage of EGR increases the particle number concentration
decreases signicantly, a reduction in number concentration of up
to 1.9  108 1/cm3 for diesel and up to 5.3  107 1/cm3 for RME
was achieved by operating the engine with EGR. The injection timing
did not have any signicant inuence on cumulative particle num-
ber concentration in the nucleation mode under EGR operation. In
the accumulation mode, the cumulative number concentration of
particle increases signicantly with increasing percentages of EGR
(Fig. 6d). At the maximum level of EGR operation, the cumulative
particle number concentration increased up to 1.7  107 1/cm3
for diesel (at 40% EGR) and up to 3.7  107 1/cm3 for RME (at 70%
EGR) when compared to the case of 0% EGR. The cumulative number
concentration of particles in the accumulation mode in Fig. 6d is
230 L. Labecki et al. / Fuel 112 (2013) 224235

0% 15% 30% 50% 70% 0% 15% 30% 40% 0% 15% 30% 50% 70% 0% 15% 30% 40% 0% 15% 30% 40% 65% 0% 15% 30% 40%
EGR EGR EGR

Fig. 6c. Cumulative number concentration of particles emitted at low load engine operation for diesel and RME in the nucleation mode at different levels of EGR and at
different injection timing.

0% 15% 30% 50% 70% 0% 15% 30% 40% 0% 15% 30% 50% 70% 0% 15% 30% 40% 0% 15% 30% 40% 65% 0% 15% 30% 40%
EGR EGR EGR

Fig. 6d. Cumulative number concentration of particles emitted at low load engine operation for diesel and RME in the accumulation mode at different levels of EGR and at
different injection timing.

lower compared to the nucleation mode discussed in Fig. 6c. It is
interesting to note that under low load operation the maximum
amount of EGR rate achieved for RME is much higher than that of
diesel fuel. EGR reduces the amount of fresh air delivered into the
combustion chamber, but the amount of fuel injected under low load
operation is less and the mixture is more premixed. So the presence
of fuel-bound oxygen in the fuel further enhances the mixing pro-
cess and improves the relative combustion performance of RME,
which was not feasible with diesel.
Generally it is known that higher injection pressure results in
lower exhaust soot emissions on mass basis. This trend could be
linked to the observations made from the particle number size dis-
tributions, particularly to the accumulation mode. From another
point of view the addition of EGR corresponds to a higher exhaust
soot formation on mass basis. The EGR effect on cumulative parti-
cles number concentration is nearly the same for both fuels. For
high levels of EGR, particles in the nucleation mode seem to coag-
ulate and tend to form larger size aggregates, which are measured
 L. Labecki et al. / Fuel 112 (2013) 224235
Fig. 7. The inuence of injection pressure on soot particles size distributions for
diesel and RME at high load engine operation.
mostly in the accumulation mode. Under low load operation RME
produced lowest cumulative particle number concentration at high
injection pressures and retarded injection timings.
3.3. Particles number size distribution (high load)
Fig. 7 shows the inuence of injection pressure on the soot par-
ticles number size distributions under high load engine operation.
A bi-modal size distribution was observed for the case of 1000 bar
and 1200 bar injection pressures for both fuels. Under high engine
load operation it can be seen that the exhaust soot particle number
concentration in the nucleation mode is lower when compared to
low engine load operation. In addition to this the accumulation
mode tends to have slightly higher particle number concentration
and the size ranges of measured particles have been shifted to-
wards larger diameters. At higher engine loads the exhaust soot
particle number concentration in the nucleation mode increases
while the number concentration in the accumulation mode de-
creases with an increase in the injection pressure for both fuels,
similar observations have been discussed in [20]. Generally it
was observed that the exhaust soot particle number size distribu-
tion for RME was lower compared to that of diesel. The differences
in size distribution at low and high engine loads are attributed to
differences in the in-cylinder fuelair mixture distribution, in-
cylinder pressure, temperature and the overall combustion pro-
cess. Under high load operation more fuel is consumed and the
whole combustion process takes place under richer mixture condi-
tions and the combustion temperatures are higher and this result
in more of a carbonaceous particles. Under low load operation
the mixtures are lean and the combustion temperatures are
Fig. 8. Inuence of EGR on soot particles size distributions for diesel and RME at
high load engine operation.
231
Fig. 9. Inuence of injection timing on soot particles size distributions for diesel
and RME at high load engine operation.
relatively low and this result is more of volatile fractions and less
carbonaceous particles. Therefore, fewer particles are seen in the
nucleation mode but more in the accumulation mode under high
engine loads when compared to low load engine operation. High
injection pressure improves air entrainment in the fuel spray and
improves mixing process leading to lower soot emissions. At
800 bar injection pressure, the soot emissions are relatively high
and this is reected in an unimodal size distribution with the high-
est number concentration in the accumulation mode for both fuels.
Increasing the injection pressure to 1000 bar and then to 1200 bar
caused a reduction in soot emission and this leads to a decreases in
the accumulation mode concentration.
Fig. 8 presents the inuence of EGR on the soot particle number
size distributions for RME and diesel at an engine load of 5 bar
BMEP, and this results in a relatively higher soot emissions com-
pared to the lower engine load. Additionally, increasing the level
of EGR further increases the soot emissions. Therefore, the maxi-
mum detectable size range of particles at the highest EGR percent-
ages (36% for diesel and 33% for RME) is larger (about 6.5682 nm
for diesel and between 38.5 nm and 682 nm for RME) than in the
case of low engine load operation, where the particle size ranges
are usually around 6.5151 nm for both fuels.
In Fig. 8, it can be seen that the size distribution up to 40 nm
was not affected by different levels of EGR but in the accumulation
mode, an increase in the number concentration of larger size par-
ticles were observed with EGR. The soot particle number concen-
tration for RME was always lower when compared to diesel. The
use of high levels of EGR caused the coagulation and aggregation
of soot particles, which was reected by the presence of high con-
centration of large particles in the accumulation mode and low
concentration of particles in the nucleation mode. Lower soot
emissions from RME can be attributed to its higher oxygen content
in the elementary composition of the fuel, when compared to die-
sel. Generally, the formation of soot is strongly dependent on the
formation of soot precursors in the rich reaction zones during com-
bustion. The soot precursors that are formed in an oxygenated fuel
react favourable with elementary oxygen to form less amounts of
aromatic compounds and soot. Additionally, very low or almost
zero sulphur content in RME compared to diesel contributes to
lower soot particle number concentration, especially in the accu-
mulation mode.
Fig. 9 shows the inuence of fuel injection timing on soot parti-
cles size distribution for the case of 1000 bar fuel injection pres-
sure. It can be seen that the nucleation mode under diesel fuel
operation was inuenced by varying the fuel injection timing un-
der high load operation. Nucleation mode was pronounced at the
fuel injection timing of 9 deg bTDC, and it was reduced when the
fuel injection timing was retarded from 9 deg bTDC to 6 deg bTDC.
232 L. Labecki et al. / Fuel 112 (2013) 224235

800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200
inj. pressure, bar inj. pressure, bar inj. pressure, bar

Fig. 10a. Cumulative number concentration of particles emitted at high engine load operation for diesel and RME fuels in the nucleation mode at different injection pressure
and injection timing.

800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200 800 1000 1200
inj. pressure, bar inj. pressure, bar inj. pressure, bar

Fig. 10b. Cumulative number concentration of particles emitted at high engine load operation for diesel and RME fuels in the accumulation mode at different injection
pressure and at different injection timing.

Retarding the fuel injection timing further to 3 deg bTDC causes
the bi-modal distribution to be transformed to an unimodal distri-
bution without any hump caused by the absence of pronounced
nucleation mode, similar observation of nuclear mode decreasing
with retarded injection time was presented in [21], where a wide
range of injection timings have been investigated. Under this load
condition the accumulation mode remained the same for diesel un-
der all injection timings. In the case of RME, the injection timing
inuenced both the nucleation and accumulation modes. Retarding
the fuel injection timing reduced the particle number concentra-
tion both in the nucleation and accumulation mode and his is asso-
ciated with the global in-cylinder temperature and mixture
composition. A signicant reduction in accumulation mode was
observed for the fuel injection timing of 3 deg bTDC for RME.
Retarding the fuel injection timing showed a reduction in the
nucleation mode in contrast to other parameters (injection
 L. Labecki et al. / Fuel 112 (2013) 224235 233

0% 15% 30% 33% 0% 15% 30% 36% 0% 15% 25% 33% 0% 15% 25% 32% 0% 15% 25% 29% 0% 15% 25% 29%
EGR EGR EGR

Fig. 10c. Cumulative number concentration of particles emitted at high load engine operation for diesel and RME in the nucleation mode at different levels of EGR and at
different injection timing.

0% 15% 30% 33% 0% 15% 30% 36% 0% 15% 25% 33% 0% 15% 25% 32% 0% 15% 25% 29% 0% 15% 25% 29%
EGR EGR EGR

Fig. 10d. Cumulative number concentration of particles emitted at high load engine operation for diesel and RME in the accumulation mode at different levels of EGR and at
different injection timing.

pressure and EGR) however the particle number in the accumula-
tion mode remained constant for diesel but decreased for RME un-
der higher engine load operation.
3.4. Cumulative particle number concentration (high load)
Figs. 10a10d show the cumulative number concentration of
particles in the nucleation and accumulation mode for diesel and
RME under different engine operating conditions. The average value
of cumulative number concentration of particles in the nucleation
mode at different injection pressure is about 1  107 (1/cm3)
for both fuels (Fig. 10a). The cumulative number concentration of
particles in the nucleation mode decreases by retarding the injec-
tion timing from 9 deg bTDC to 3 deg bTDC for both diesel and
RME. However, the differences in the cumulative number concen-
tration between injection timings were small for RME compare to
diesel. No clear trends could be seen for the effect of fuel injection
pressures on the cumulative number concentration of particles in
234 L. Labecki et al. / Fuel 112 (2013) 224235
the nucleation mode for RME. But the cumulative number concen-
tration for diesel in the nucleation mode show a marginally increas-
ing trend with an increase in injection pressure, particularly at an
injection timing of 9 deg bTDC for diesel. The cumulative number
concentration of particles in the nucleation mode are much lower
at higher engine loads compared to low load operation, almost an
order of magnitude difference was observed between these two
loads. The number concentration of particles in the nucleation
mode is lower for RME compared to diesel under high load
operation. Fig. 10b shows the effect of injection pressure on the
cumulative number concentration of particles emitted in the accu-
mulation mode. The cumulative number concentration decreases
with higher injection pressure, on an average, the cumulative num-
ber concentration reduced from 1.0  107 (1/cm3) to 1.2  106
(1/cm3). The cumulative number concentration of particles emitted
in the accumulation mode by RME is lower than diesel for a given
fuel injection timing. Thus the trend of observing larger size
particles in the accumulation mode may correspond to the smoke
number measurements. This also demonstrates the fact that soot
emissions can be reduced by increasing the injection pressure due
to better entrainment and mixing of fuel with air as reported in
[22].
The effect of EGR on the cumulative soot particle number con-
centration in the nucleation mode is shown in Fig. 10c. It can be
seen that EGR did not have any signicant inuence on cumulative
particle number concentration for diesel, it is almost constant for
different levels of EGR, and it is about 6.5  106 (1/cm3). Similar
observation was made for RME under all conditions of EGR except
for the case of maximum EGR. At the highest level of EGR opera-
tion, the cumulative particle number concentration signicantly
decreases for RME compared to diesel. No clear trends were ob-
served for the effect of injection timing on the cumulative particle
number concentration for both fuels. The cumulative particle num-
ber concentration of particles in the nucleation mode at different
levels of EGR is relatively less at high load when compared to
low engine load operation. On the other hand, it can be seen in
Fig. 10d that the cumulative particle number concentration in
the accumulation mode increases signicantly with increasing lev-
els of EGR for both fuels. For a given injection timing, RME showed
lower cumulative particle number concentration than diesel for
most of the operating conditions. The particles in the nucleation
size range agglomerated and formed larger diameters particles as
well as aggregates, which shifted its distribution from nucleation
mode to accumulation mode for higher percentages of EGR. Injec-
tion timing did not have any signicant effect on the cumulative
particle number concentration for both fuels. Under high load
operation the global in-cylinder temperatures are high, together
with EGR operation the rich mixture results in higher rate of soot
formation but the presence of fuel-bound oxygen tends to reduce
the soot formation rate but this effect was not that dominant as
it was observed under the low load operation thus the particle
number concentration in the accumulation mode is high.
4. Conclusions
Particles number size distributions from the combustion of die-
sel and RME fuels in a HSDI engines were investigated under differ-
ent engine operating conditions such as fuel injection timing,
injection pressure, EGR and engine load. Some of the main ndings
are summarised as follows:
 At low engine load, more particles were observed in the nucle-
ation mode. This condition provides low soot emissions due to
lean mixture combustion. The nucleation mode particles were
observed as a result of limited condensation and agglomeration.
Increasing the engine load to 5 bar BMEP caused the combus-
tion mixture to become richer and favourable for higher soot
emissions. All small size particles tend to coagulate and form
aggregates of larger size particles, therefore the accumulation
mode, in the case of higher engine load, had wider particle size
range.
 Increasing the fuel injection pressure reduced the particle size
and its concentration in the accumulation mode under low load
operation. At higher loads, the nucleation mode increases as the
fuel injection pressure increases, while the number concentra-
tion in the accumulation mode decreases with an increase in
injection pressure.
 The addition of EGR caused smaller particles from the nucle-
ation mode to agglomerate to form larger size particles, which
were observed in the accumulation mode. This effect was
observed for EGR at both low and high loads. A combination
of high load and EGR resulted in the detection of signicantly
larger size particles. High levels of EGR caused the particle num-
ber size distribution to become uni-modal. The number concen-
tration in the accumulation mode under high load operation
increased with EGR.
 The effect of injection timing did not have a clear and consistent
trend between diesel and RME fuels in the nucleation and accu-
mulation mode under low load operation. Under high load
operation, retarding the fuel injection timing from 9 deg bTDC
to 3 deg bTDC caused the particle number concentration in
the nucleation mode to decrease for diesel. The size distribution
transformed from bi-modal to uni-modal distribution at the
injection timing of 3 deg bTDC, but the accumulation mode
remained the same for different injection timings. The particle
number concentration for RME reduced both under nucleation
and accumulation mode when the fuel injection timing was
retarded at high load.
RME generally emitted lower soot particle concentration than
diesel under most of the operating conditions. This is mainly due
to higher oxygen content in the elementary composition of RME.
The investigation reveals that the presence of oxygen in the fuel
has the potential to lower the particle number concentration emis-
sion in diesel engines.
Acknowledgments
Authors are thankful to Andrew Selway, Kenneth Anstiss and
Clive Barrett for their continuous support and technical help dur-
ing the measurements. Authors would like to acknowledge the
support that was provided from Shell, UKIERI and EPSRC (EP/E
050085/1) Projects.
References
[1] Tree DR, Svensson KI. Prog Energy Combust Sci 2007;33:272309.
[2] Johnson T. Diesel emissions in review. SAE Int J Engines 2011;4(1):14357.
[3] Akihama K, Takatori Y, Inagaki K, Sasaki S, Dean AM. Mechanism of the
smokeless rich diesel combustion by reducing temperature. SAE technical
paper 2001-01-0655; 2001.
[4] Kimura S, Aoki O, Ogawa H, Muranaka S, Enomoto Y. New combustion concept
for ultra-clean and high-efciency small DI diesel engines. SAE technical paper
1999-01-3681; 1999.
[5] Yokota H, Kudo Y, Nakajima H, Kakegawa T, Suzuki T. A new concept for low
emissions diesel combustion. SAE technical paper 970891; 1997.
[6] Shimazaki N, Akagawa H, Tsujimura K. An experimental study of premixed
lean diesel combustion. SAE technical paper 1999-01-0181; 1999.
[7] Kittelson DB, Watts WF, Johnson JP. On-road and laboratory evaluation of
combustion aerosols - Part1: Summary of diesel engine results. Aerosol Sci
2006;37:91330.
[8] Mathis U, Ristimaki J, Mohr M, Keskinen J, Ntziachristos L, Samaras Z, et al.
Sampling conditions for the measurement of nucleation mode particles in the
exhaust of a diesel vehicle. Aerosol Sci Technol 2004;38:114960.
 L. Labecki et al. / Fuel 112 (2013) 224235
[9] Kittelson DB, Johnson J, Watts W, Wei Q, Drayton M, Paulsen D, Bukowiecki N.
Diesel Aerosol Sampling in the Atmosphere. SAE, 200-01-2212; 2000.
[10] Giechaskiel B, Ntziachristos L, Samaras Z, Scheer V, Casati R, Vogt R. Formation
potential of vehicle exhaust nucleation mode particles on-road and in the
laboratory. Atmos Environ 2005;39:31918.
[11] Vouitsis E, Ntziachristos L, Samaras Z. Theoretical investigation of the
nucleation mode formation downstream of diesel after-treatment devices.
Aerosol Air Qual Res 2008;8:3753.
[12] Burtscher H. Physical characterization of particulate emissions from diesel
engines: a review. Aerosol Sci 2005;36:896932.
[13] Lyyranen J, Jokiniemi J, Kauppinen EI, Backman U, Vesala H. Comparison of
different dilution methods for measuring diesel particle emissions. Aerosol Sci
Technol 2004;38:1223.
[14] Hueglin Ch, Scherrer L, Burtscher H. An accurate, continuously adjustable
dilution system (1:10 to 1:104) for submicron aerosols. J Aerosol
Sci:1997;28:10491055.
[15] Lyyrnen J, Jokiniemi J, Kauppinen EI, Joutsensaari J. Aerosol characterisation
in medium-speed diesel engines operating with heavy fuel oils. J Aerosol Sci
1999;30:77184.
[16] Eastwood P. Particulate emissions from vehicles. Chichester: John Willey &
Sons Ltd.; 2008.
235
[17] Winklmayr W, Reischl GP, Lindner AO, Berner A. A new electromobility
spectrometer for the measurement of aerosol size distributions in the size
range from 1 to 1000 nm. J Aerosol Sci 1991;22(3):28996.
[18] Guerrassi N, Dupraz P. A common rail injection system for high speed direct
injection diesel engines, SAE technical paper, 980803. Warrendale, PA: SAE;
1998.
[19] Tsolakis. Effects on particle size distribution from the diesel engine operating
on RME-biodiesel with EGR. Energy Fuels 2006;20(4):141824.
[20] Lhde Tero, Rnkk Topi, Happonen Matti, Sderstrm Christer, Virtanen
Annele, Solla Anu, et al. Effect of fuel injection pressure on a heavy-duty diesel
engine nonvolatile particle emission. Environ Sci Technol 2011;45(6):
25049.
[21] Benajes Jesus, Novella Ricardo, Arthozoul Simon, Kolodziej Christopher.
Particle size distribution measurements from early to late injection timing
low temperature combustion in a heavy duty diesel engine. SAE Int J Fuels
Lubri 2010;3(1):56781.
[22] Wang Xiangang, Zuohua Huang, Wu Zhang, Olawole Abiola Kuti, Keiya
Nishida. Effects of ultra-high injection pressure and micro-hole nozzle on
ame structure and soot formation of impinging diesel spray. Appl Energy
2011;88(5):16208.