Fuel 235 (2019) 288–300

Contents lists available at ScienceDirect

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

Full Length Article

Study of combustion, performance and emissions of a diesel engine fueled T

with ternary fuel in blended and fumigation modes
⁎
Meisam Ahmadi Ghadikolaeia, , Chun Shun Cheunga, Ka-Fu Yungb
a
Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong
b
Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong

A R T I C LE I N FO A B S T R A C T

Keywords: The present study aimed to compare the combustion, performance and emissions of a diesel engine fueled with
Diesel engine diesel, biodiesel and ethanol in the blended and fumigation modes. In addition, a combined fumigation and
Performance blended mode (F + B) of fueling was introduced to understand its behavior in comparison with the blended and
Emissions fumigation modes. In the blended mode, the fuel was composed of 80% diesel, 5% biodiesel and 15% ethanol
Diesel-biodiesel-ethanol (DBE) blend
(DBE), by volume. In the fumigation mode, a mixture of biodiesel and ethanol (BE) was injected into the intake
Fumigation mode
manifold and diesel fuel was used as the main fuel. In the F + B mode, half of the BE mixture was injected into
Dual-fuel operation
the intake manifold and another half of the DB mixture was mixed with diesel to form the main fuel. The
experiments were conducted at a speed of 1800 rpm with five engine loads at a constant mixture percentage of
D80B5E15 for the three fueling modes to provide the same fuel composition for comparing the effects of the
three fueling modes.
According to the average of five loads, compared to the pure diesel operation, the blended mode causes
increases in peak heat release rate (HRR), ignition delay (ID), BSFC, CO, HC and NO2; drops in duration of
combustion (DOC), CO2, NOX, PM, total number concentration (TNC) and geometric mean diameter (GMD); and
similar peak in-cylinder pressure and BTE. While the fumigation mode leads to higher peak HRR, DOC, BSFC,
CO, HC, NO2 and TNC; lower ID, BTE and NOX; and similar peak in-cylinder pressure, CO2, PM and GMD in
comparison with the diesel mode. In addition, the F + B mode has the effects between those of the fumigation
and blended modes. Among the three fueling modes, the blended mode gives the highest engine performance
and lowest emissions, except for NOX. In contrast, the fumigation mode gives the worst engine performance and
emissions, except for NOX.

1. Introduction
The use of biofuels in diesel engine is an attractive method for re-
ducing reliance on diesel fuel and for reducing engine emissions [1–3].
In particular, the diesel-biodiesel-ethanol (DBE) blend has the potential
to be an alternative to diesel fuel for diesel engines [3,4]. DBE is stable
even below 0 °C and has identical or superior fuel properties to regular
diesel fuel [5]. Also, the addition of biodiesel in the diesel-ethanol
blend shows a favorable approach towards the formulation of a novel
form of alternative fuel [6], because diesel and ethanol are immiscible.
The literature reveals that DBE has been utilized in many studies to
investigate its effects on the combustion, performance and emissions
characteristics in different types of diesel engines. The findings of these
studies have been summarized in the review paper [7]. In addition to
the blended mode, biofuels can be applied to a diesel engine in the
fumigation mode by injecting the biofuels into the air intake to form a
uniform air/fuel mixture inside the engine cylinder [8,9]. Most of the
studies concerning fumigation have been conducted by using alcohols
(especially ethanol and methanol) as the fumigated fuels and diesel as
the main fuel.
Each fueling mode (blended or fumigation) has its advantages and
disadvantages [7–9]. For instance, the blended mode causes increase in
BTE, decrease in HC and CO emissions but increase in NOx emissions in
most cases. On the other hand, the fumigation mode has reverse effects
of the blended mode, leading to drop in NOX but reduction in BTE and
increase in HC and CO emissions in most cases. Most of the former
studies were conducted either in the blended mode or in the fumigation
mode. Only a few studies [10–13] were conducted to compare the effect
of these two modes using the same fuel on the same diesel engine and
under the same operating conditions. Abu-Qudais et al. [10] compared

⁎
Corresponding author.
E-mail address: meisam.ahmadi67@yahoo.com (M.A. Ghadikolaei).

https://doi.org/10.1016/j.fuel.2018.07.089
Received 9 May 2018; Received in revised form 29 June 2018; Accepted 20 July 2018
Available online 31 July 2018
0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

Nomenclature E Ethanol
EOC End of combustion
X̄i Sample observation F+B Fumigation + blended mode
X1 , ⋯, XN Independent variables FC Fuel consumption
Xm Mean of measurements FG Fumigated gasoline
ω1, ⋯, ωN Uncertainties of independent variables GMD Geometric mean diameter
ωR Uncertainty of the result H Hydrogen
ωr Random error HCLD Heated chemiluminescent analyzer
ωs Systematic error HFID Heated flame ionization detector
°CA Crank angle degree HRR Heat release rate
B Biodiesel Hz Hertz
B20 20% biodiesel + 80% diesel ID Ignition delay
B20 + FG (20% biodiesel + 80% diesel) + fumigated gasoline min Minute
BE Biodiesel-ethanol N Particle number
BSCO Brake specific carbon monoxide NDIR Non-dispersive infrared analyzers
BSCO2 Brake specific carbon dioxide Nm Newton meter
BSFC Brake specific fuel consumption nm Nanometer
BSHC Brake specific hydrocarbon O Oxygen
BSNO2 Brake specific nitrogen dioxide pC Pico coulomb
BSNOX Brake specific nitrogen oxides R Function
BSPM Brake specific particulate matter rpm Revolutions per minute
BTDC Before top dead center SMPS Scanning mobility particle sizer
BTE Brake thermal efficiency SOC Start of combustion
C Carbon TEOM Tapered element oscillating microbalance
CI Compression ignition THC Total hydrocarbon
CPC Condensation particle counter TNC Total number concentration
D Diesel ULSD Ultra-low-sulfur diesel
D + FG Diesel + fumigated gasoline N Number of samples
DBE Diesel- biodiesel- ethanol SD Standard deviation
DI Direct injection U Total uncertainty
DMA Differential mobility analyzer t Student's t statistic
DOC Duration of combustion ν Degrees of freedom
DR Dilution ratio

the effects of ethanol fumigation and ethanol-diesel blends on the
performance and emissions of a single cylinder diesel engine under
different engine speeds, using 20% of ethanol blended with diesel and
20% of ethanol (ethanol percentage as a fraction of the diesel energy
input at full rack setting) in the fumigation mode. They found im-
provement in BTE and soot emission but increase in HC and CO emis-
sions in both modes of operation compared to the operation with pure
diesel fuel; while the fumigation mode showed higher BTE and lower
CO, HC and soot emissions, in comparison with the blended mode.
However, their investigation did not involve NOx and PM emissions.
Cheng et al. [11] compared the influences of 10% blended methanol
(by volume) or 10% fumigation methanol with waste cooking oil bio-
diesel on the performance and emissions of a diesel engine at a constant
engine speed of 1800 rpm with five engine loads. Their results in-
dicated, compared to the diesel fuel operation, there was a reduction of
CO2, NOX, particulate mass emissions and mean particle diameter in
both modes of operation; while CO and HC emissions were similar in
the blended mode but higher in the fumigation mode. Compared to the
blended mode, the fumigation mode gave slightly higher BTE at
medium and high engine loads and lower NOx emission, but there was
also higher CO, HC and NO2 and particulate emissions. Şahin et al. [12]
compared the impact of n-butanol/diesel blends and n-butanol fumi-
gation on the performance and emissions of a turbocharged diesel en-
gine, using 2, 4, and 6% by volume of n-butanol, under different engine
loads and engine speeds. It was found that both modes led to reduction
in smoke compared to the diesel fuel while the reduction was higher in
the fumigation mode. Compared to the diesel fuel, NOX emission was
lower in the fumigation mode but higher in the blended mode; HC
emissions were higher in both modes, being higher in the fumigation
mode than the blended mode; and similarly, CO2 emissions were
increased for the two modes, being lower in the fumigation mode than
the blended mode. In regard to BSFC, fumigation caused rise in BSFC at
all test conditions, while the blended mode led to reduction in BSFC for
2% and 4% n-butanol blends (6% n-butanol blend caused increase in
BSFC). Also, the HRR diagrams of the blended mode were similar to
those of diesel fuel; however, those of the fumigation mode had a
double peak structure. The first peak, which was small, occurred earlier
than that of diesel fuel and the second peak, which was also the main
peak, took place later. Their investigation did not involve PM emissions.
Also, Mariasiu et al. [13] studied the performance and emissions of a
single cylinder diesel engine using a diesel fuel containing 5% by vo-
lume of rapeseed biodiesel, a blended fuel containing 85% of the diesel
fuel and 15% by volume of ethanol and the diesel fuel with fumigated
ethanol under a constant speed of 1800 rpm and different engine loads.
It was found that the fumigation mode had lower BSFC, NOX and THC,
and higher CO emissions compared to those of the blended mode.
However, their investigation did not involve PM emissions.
According to the literature, only alcohol fuels were utilized in the
fumigation mode and the use of biodiesel + alcohol as a fumigated fuel
almost cannot be found. Since DBE is a widely investigated blended
fuel, there is an interest to apply it in the fumigation mode to identify its
advantages and disadvantages in comparison with the blended mode.
Therefore, this study aims at covering the above knowledge gaps by
conducting experiments to compare the effect of using blended mode
(DBE as the fuel) with fumigation mode (BE as the fumigated fuel and
diesel as the main fuel to form DBE) on the engine combustion, per-
formance and emissions. Moreover, an additional mode of operation,
using 50% of the BE blended with the diesel fuel and applying another
50% of the BE mixture in the fumigation mode (namely F + B mode), is
also introduced for comparison with the other two modes. According to

289
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

the authors' knowledge, this investigation is novel because it involves Table 1

comparing the blended mode with fumigation mode using DBE fuels Specifications of tested engine.
and a new mode of operation, the F + B mode, is also introduced in this Model Isuzu 4HF1
study.
Engine type In-line 4-cylinder DI
Combustion chamber shape Omega
2. Experimental setup and procedure Maximum power 88 kW/3200 rpm
Maximum torque 285 Nm/1800 rpm
Bore × stroke 112 mm × 110 mm
The experiments were performed on a 4-cylinder direct-injection,
Displacement 4334/cc
water-cooled diesel engine. Similar engines for small trucks or non-road Compression ratio 19.0: 1
applications are still used in Hong Kong and China. The engine speed Fuel injection timing 8° BTDC
and torque were controlled by an eddy-current dynamometer and Ono Injection pump type Bosch in-line type
Sokki heavy diesel engine test system. The schematic of the experi- Injection nozzle Hole type (with 5 orifices)

mental setup and the specifications of the engine are shown in Fig. 1
and Table 1, respectively. Fig. 1 shows that the fumigated fuel is in-
Table 2
jected into the air intake with fuel injectors, one for each cylinder, with
Properties of the tested fuels [15].
an injection pressure of 0.35 MPa.
Experiments were conducted at a constant engine speed of Parameters Diesel Biodiesel Ethanol
1800 rpm with five engine torques of 57, 99.8, 142.5, 185.3 and
Cetane number 52 51 6
228 Nm, corresponding to 20, 35, 50, 65 and 80% of the full engine Lower heating value (MJ/kg) 42.5 37.5 28.4
torque. At each operating condition, the engine was allowed to run for Density (kg/m3) at 20 °C 840 871 786
around one hour until the exhaust gas temperature, the cooling water Viscosity (mPa S) at 40 °C 2.4 4.6 1.2
temperature, the lubricating oil temperature as well as the CO2 con- Heat of evaporation (kJ/kg) 250–290 300 840
Carbon content (% mass) 86.6 77.1 52.2
centration reached steady-state conditions and data were measured Hydrogen content (% mass) 13.4 12.1 13
subsequently. Oxygen content (% mass) 0 10.8 34.8
The chosen fuels are ultra-low-sulfur diesel (ULSD) with fuel sulfur Sulfur content (% mass) less than10 less than10 0
content of less than 10 ppm by weight, biodiesel and ethanol (high Flash point (°C) 78 210 15
Stoichiometric air–fuel ratio a 14.56 12.59 9.00
purity of 99.9%). The biodiesel was produced from waste cooking oils
collected from restaurants in Hong Kong by Dynamic Progress and the a = calculated according to [16].
biodiesel qualities were in compliance with EN14214 standard.
Composition of the biodiesel is shown in the literature from the same involves the use of pure diesel (ultra-low-sulfur diesel, ULSD) to obtain
laboratory [14], comprising 11.46% methyl palmitate, 4.22% methyl the baseline results. The second mode is the blended mode, in which the
stearate, 35.22% methyl oleate, 39.73% methyl linoleate, 6.24% me- ternary blend contains 80% diesel, 5% biodiesel and 15% ethanol by
thyl linolenate and a small amount of some other fatty acid methyl volume. The third mode is the fumigation mode. In this case, BE mix-
esters. In addition, the properties of the tested fuels are presented in ture comprising 25% biodiesel and 75% ethanol is applied in the fu-
Table 2. migation mode while diesel fuel is injected into the engine using the
Four fueling modes are involved in this study. The first mode

Fig. 1. Schematic diagram of the experimental setup.

290
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

original fuel injectors, with a desired overall fuel consumption com-
prising 80% diesel and 20% BE (5% biodiesel and 15% ethanol), to
match with the composition of the fuel used in the blended mode. An
electronic control unit (ECU) was used to adjust and control the fueling
rate of BE while the amount of diesel fuel was adjusted automatically
through the Ono Sokki heavy diesel engine test system. Two electronic
balances were used to measure the mass consumption rates of ULSD and
BE. To achieve the 20% of fumigated fuel, firstly the engine was run by
diesel only at the desired engine load, then the fumigated fuel (BE) was
injected into the cylinder (the mass of diesel fuel would be reduced
automatically by the Ono Sokki heavy diesel engine test system to
maintain the desired engine load) and the consumed masses of diesel
and fumigated fuel were recorded with the two electronic balances. If
the consumed volume of the fumigated fuel was less than 20% of the
total volume (diesel + fumigated fuel), the amount of injected BE fuel
was increased by adjusting the ECU. In contrast, if the consumed vo-
lume of fumigated fuel was more than 20% of the total volume, the rate
of injected BE fuel was decreased by adjusting the ECU. After several
trials on the measurements of ULSD and BE mass flow rates, the desired
fuel content (D80B5E15) for each engine load could be achieved. The
fourth mode is the F + B mode. In this mode, half of the BE mixture was
injected into the intake manifold (as fumigation mode) while another
half of the BE mixture was blended with ULSD to form a DBE mixture
comprising 90% diesel, 2.5% biodiesel and 7.5% ethanol for direct in-
jection into the diesel engine. Therefore, the final fuel content in the
F + B mode is also D80B5E15.
In this study, the in-cylinder pressure was measured using a piezo-
electric pressure sensor (type 6056A, Kistler Co., Inc.) and a charge
amplifier (type 5011B, Kistler Co., Inc.). The crank-angle signal was
measured by a crank-angle encoder (type CA-RIE-360/720, DEWET-
RON GmbH) with resolution of 0.5°CA. First Law of Thermodynamics
was employed to convert the pressure data to HRR data using the
commercial software DEWESoftTM (DEWETRON GmbH). Regarding the
gaseous emissions, NOX was measured with a heated chemiluminescent
analyzer (600 HCLD, CAI Inc.), HC was measured with a heated flame
ionization detector (300 HFID, CAI Inc.), CO and CO2 were measured
with non-dispersive infrared analyzers (300 NDIR, CAI Inc.). All the
analyzers were calibrated before the measurements. All gaseous emis-
sions were recorded over a period of five minutes to obtain the average
values. The fuel consumptions in the blended and fumigation modes
were recorded gravimetrically separately by two electronic balances
(Shimadzu Balance, Model BX-32KH) with readability and precision of
0.1 g.
In respect of PM sampling, the exhaust gas was diluted with a two-
stage mini-diluter (Dekati Ltd, Finland). The sampled gas was taken
from the exhaust manifold through an insulated and heated sampling
line (at 155 °C) to prevent the condensation of volatile substances and
deposition of solid particles on the interior pipe wall. The first stage was
heated with an electrical heater while the second stage was not. The
actual dilution ratio in this study was calculated according to Eq. (1).
[CO2 ] e−[CO2 ] b
DR =
[CO2 ] d−[CO2 ] b (1)
where [CO2]e, [CO2]b and [CO2]d are the CO2 concentrations of the
raw exhaust gas, background air and diluted exhaust, respectively. In
the present study, depending on the engine operating condition, the
first-stage DR varied from 6.0 to 8.3 and the second-stage DR varied
from 49.0 to 67.9. The first-stage diluted exhaust gas was delivered to a
tapered element oscillating microbalance (TEOM 1105, Rupprecht &
Patashnick Co., Inc.) with resolution of 0.1mg / m3 for measuring parti-
culate mass concentration, and the second-stage one was connected to a
scanning mobility particle sizer (SMPS model 3494, TSI Inc.) for mea-
suring the particle size distribution and number concentration, for
particles in the range of 15–750 nm. The SMPS consists of a differential
mobility analyzer (DMA model 3071 A, TSI Inc.) and a condensation
particle counter (CPC model 3022, TSI Inc.). PM mass concentration
was recorded continuously for five minutes while particle number-size
distribution was measured four times at each operating condition and
the averages of the results are presented.
2.1. Error analysis and uncertainty analysis
The total (overall) uncertainty in the present study was calculated
by the root-sum-square combination of fixed and random errors con-
tributions, according to Moffat [17], as shown below:
Xi = X¯i ± U (2)
U= (ωs )2 + (ωr )2 (3)
where X̄i is the observation in a single-sample experiment or the mean
of a set of N observations in a multiple-sample experiment; U is the
total uncertainty, ωs is systematic error and ωr is random error.

Table 3
Type of equipment with range and accuracy.
Parameters Type of equipment Range Accuracy Uncertainty % (at 95%)

NO or NOX 600 HCLD Analyzer 0–3000 ppm (NO or NOX) ± 0.5% ± 1.00
HC 300 HFID Analyzer 0–30000 ppm of carbon ± 1% ± 1.98
CO2 300 NDIR Analyzer 0–20% ± 1% ± 1.96
CO 300 NDIR Analyzer 0–25000 ppm ± 1% ± 1.97
PM Mass TEOM 1105 0.1mg / m3 to several g / m3 ± 0.1mg / m3 ± 1.91
TNC and GMD SMPS model 3494 5–1000 nm N ± 2.07 TNC
±
N ± 0.87 GMD
Rotational Speed FC 3000 Engine Automatic Measure and 0–8000 rpm ± 1% Accuracy and ± 5 rpm for
Control System Control
Torque FC 3000 Engine Automatic Measure and 0–600 Nm ± 0.3% Accuracy and ± 0.2% for –
Control System Control
Fuel Mass Consumption BX32KH Digital Balance 0–32 kg ± 0.2 g –
Conversion of Pressure Signal to Charge Amplifier, Kistler Co., Type 5011B ± 10 - ± 999000 pC ± 0.5% –
Voltage
In-Cylinder Pressure Piezoelectric Pressure Sensor, Kistler Co., 0–250 bar ≤ ± 0.3% –
Type 6056A
Crank-Angle Signal Acquisition CA-RIE-360/720 Encoder, (DEWETRON 0–6000 rpm ± 2.5% –
GmbH)
BSFC ± 2.15
BTE ± 2.16
ID or DOC ± 1.95
Peak Pressure ± 1.16
Peak HRR ± 1.98

291
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

Systematic error is the accuracy of instrument based on the manu- where ωR is the uncertainty of R and ω1,ω2,ω3 , ⋯, ωN are the un-
facturers' specifications (Table 3) and random error is obtained from: certainties of the independent variables.
t × SD The uncertainty in the result can be determined through Eq. (8).
ωr =
√N (4) 2 2 2 2 1/2
ωR = ⎡ ⎛ ∂R ⎞ ⎛ ∂R ⎞ ⎛ ∂R ⎞ ⎛ ∂R ⎞⎤
⎢ ∂X1 ω1 + ∂X2 ω2 + ∂X3 ω3 +…+ ∂XN ωN ⎥
⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟

where t represents the Student's t statistic appropriate for the number of ⎣⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠⎦ (8)
samples (N ) and the confidence level desired (t = 1.96 for 95% con-
In the present study, the uncertainties were computed with 95%
fidence level), and SD is the standard deviation of the set of N ob-
confidence level with N = 300 for the gaseous emissions and N = 500
servations.
for combustion analysis.
The standard deviation of the large observations (300 for emission
The steady state experiments were repeated three times for ensuring
and 500 for combustion analysis in this study), is defined by:
that the data were repeatable within the experimental uncertainties of
1 2
N the measurements. The experimental results were compared with the
⎡1
SD = ⎢
N
∑ (Xi −Xm )2⎤⎥ two-tailed Student’s T-test to verify if the results are significantly dif-
⎣ i=1 ⎦ (5) ferent from each other at 95% significance level. The equipment ac-
where N is the number of measurements, X̄i is the observation of each curacies and experimental uncertainties in the measurements are pre-
sample and Xm is the arithmetic mean of measurements calculated by: sented in Table 3.
n
1 3. Results and discussion
Xm =
N
∑ X¯i
i=1 (6)
The combustion processes of the four fueling modes are similar to
The above Eqs. ((2)–(6)) are used for determining the uncertainties
each other, consisting of a premixed combustion phase followed by a
of data measured directly by the instruments and analyzers. For para-
diffusion combustion phase. However, the magnitudes of these com-
meters that cannot be measured directly (e.g. torque, power, BSFC and
bustion phases and other combustion parameters (e.g. HRR, in-cylinder
BTE), their uncertainties can be estimated from the uncertainties in the
pressure, ID and DOC) varied for each fueling mode due to different fuel
primary measurements, using Eqs. (7)–(8).
properties and modes of ignition. In diesel mode, pure diesel fuel is
If the result R is a given function of the independent variables
directly injected into the engine cylinder and the ignition is initiated by
X1 , X2 , X3 , ⋯, XN , then:
hot compressed air. In the blended mode, diesel, biodiesel and ethanol
R = R (X1 , X2 , X3 , ⋯, XN (7) are uniformly mixed, directly injected into the engine cylinder and

a) 70 Diesel (57 Nm)

160
b)
70
Diesel (142.5 Nm)
160

Blend (57 Nm) Blend (142.5 Nm)

60 140 60 140
F+B (57 Nm) F+B (142.5 Nm)
Fumigation (57 Nm) Fumigation (142.5 Nm)
120 120
Heat release rate (J/°CA)
In-cylinder pressure (Bar)

Heat release rate (J/°CA)

In-cylinder pressure (Bar)

50 50
100 100
40 40
80 80
30 30
60 60

20 40 20 40

10 20 10 20

0 0
0 0
-10 0 10 20 30 40 -10 0 10 20 30 40
Crank angle (°CA) Crank angle (°CA)

c) 70 Diesel (228 Nm)

Blend (228 Nm)
160

60 F+B (228 Nm) 140

Fumigation (228 Nm)

120
Heat release rate (J/°CA)
In-cylinder pressure (Bar)

50
100
40
80
30
60

20 40

10 20

0
0
-10 0 10 20 30 40
Crank angle (°CA)
Fig. 2. Variations of in-cylinder pressure and heat release rate with (a) low, (b) medium and (c) high loads.

292
M.A. Ghadikolaei et al.
ignited by hot compressed air. However, in the fumigation mode,
mixture of biodiesel and ethanol (BE) is injected into the engine cy-
linder through the intake port and the main fuel (diesel) is directly
injected into the engine. During the intake and compression strokes, the
BE is mixed with air to form a homogeneous mixture and then the main
fuel is injected into the cylinder. The main fuel is ignited by hot com-
pressed air while the fumigated fuel is ignited by the main fuel. In the
F + B mode, the main fuel is a mixture of diesel, biodiesel and ethanol,
rather than pure diesel.
Since initiation and quality of combustion are dependent on the
fuel, oxidizer (air) and temperature (also chain reaction to sustain the
flame), the combustion parameters, performance and emissions are
almost different for the four fueling modes. In the blended mode, the
lower cetane number and higher heat of evaporation of ethanol cause
late combustion and affect other combustion parameters compared to
the pure diesel mode. In the fumigation mode, the combustion tem-
perature is low due to the fumigated ethanol; however the homo-
geneous fumigated fuel/air mixture and using diesel as main fuel which
has higher cetane number might lead to different effects on the com-
bustion parameters, engine performance and emissions. Therefore, the
combustion, performance and emissions of the engine are varied with
different fueling modes, as shown in the results presented in the fol-
lowing sections.
3.1. Engine combustion
The combustion characteristics in the present study, including the
HRR, were obtained based on the average of 500 cycles of in-cylinder
pressure to minimize the influence of cycle-to-cycle variations. Start of
combustion (SOC) is defined as the beginning of heat release. Ignition
delay (ID) means the interval between start of fuel injection (constant at
8 °CA BTDC for the used engine) and SOC. End of combustion (EOC) is
the crank angle of 95% heat release. Duration of combustion (DOC) is
the interval between SOC and EOC.
Fig. 2 illustrates the typical curves of in-cylinder pressure and HRR
at 1800 rpm with low (57 Nm), medium (142.5 Nm) and high (228 Nm)
engine loads. The peak in-cylinder pressure and peak HRR values are
shown in Fig. 3(a) and (b), respectively. Figs. 2 and 3(a) reveal that the
in-cylinder pressure increases, and occurs further away from the top
dead center, with increase in engine load for all the tested fueling
modes. Similar trend was recorded in the literature using diesel,
blended [15,18,19] and fumigation modes [20]. However, Figs. 2 and
3(b) show that the peak HRR increases with rise in load from 57 Nm to
70 Diesel F+B
a) Blend Fumigation
60
Peak in-cylinder pressure (Bar)
50
40
30
20
10
0 0
57 99.8 142.5
Load (Nm)
Fig. 3. Variations of (a) peak in-cylinder pressure and (b) peak HRR with engine load.
Fuel 235 (2019) 288–300
185.3 Nm and then decreases at 228 Nm for all the fueling modes. Si-
milar trend was observed in the literature for diesel, blended [15,18,21]
and fumigation modes [20].
Figs. 2 and 3(a) also shows that, compared to the diesel fuel op-
eration, the peak in-cylinder pressures at the other three modes of op-
eration are lower at low engine load, similar at medium load and higher
at high load. The lower peak in-cylinder pressures at low engine load is
due to the lower combustion temperatures and the longer ignition de-
lays (Fig. 4(a)) which lead to SOC occurring further away from the top
dead center [15,18]. On the other hand, the higher peak in-cylinder
pressures at high engine load is due to the higher combustion tem-
peratures and shorter ignition delays of the F + B and fumigation
modes (Fig. 4(a)). Some other studies also found that the peak in-cy-
linder pressure is lower at low engine load and higher at high engine
load in blended mode [15,18] and fumigation mode [20,22] in com-
parison with the baseline fuels. However, results based on the average
of five loads indicate that the differences between the peak in-cylinder
pressures of the blended, F + B and fumigation modes with diesel fuel
operation are −0.2%, 0.9% and 1.3%, respectively, which are not
statistically significant at 95% confidence level.
For HRR, the blended, F + B and fumigation modes have higher
peak HRR at almost all the tested engine loads (except for fumigation
mode at 57 Nm) in comparison with the pure diesel, with increase of
20.4, 6.3 and 2.2% in the blended, F + B and fumigation modes of
operation, respectively, on the average of five loads. For the blended
mode, this increase in peak HRR can be attributed to the better vola-
tility and lower viscosity (better fuel atomization) of ethanol, and
longer ignition delay which lead to accumulation of more fuel during
the ignition delay time to burn in the premixed burning phase [23–25].
In the fumigation mode, the ignition delay is shorter (Fig. 4(a)), the
duration of combustion is longer (Fig. 4(b)), the fumigated fuel is not
well atomized (due to low fuel injection pressure of 0.35 MPa) and
fumigated fuel mixing with air is poor, compared to the blended mode;
therefore the fumigation mode has less effect on the increase in peak
HRR (only 2.2%). In addition, the lean mixture of fumigated fuel and
air plus the low combustion temperature at the low load of 57 Nm cause
deterioration in the combustion efficiency [11] and even a drop in the
peak HRR. The change in peak-HRR in the F + B mode is between those
of the blended and fumigation modes.
Fig. 4 (a) and 4 (b) show that the increase in engine load causes
reduction in ignition delay (ID) and increase in duration of combustion
(DOC) for all the tested fueling modes. The ID is reduced due to increase
in in-cylinder gas temperature [6] as a consequence of increase in en-
Diesel
b) Blend
100 F+B
Fumigation
Peak heat release rate (J/°CA)
80
60
40
20
185.3 228 57 99.8 142.5 185.3 228
Load (Nm)
293
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

Diesel F+B Diesel

14 Blend Fumigation b) 60 Blend
a) F+B
Fumigation

Duration of combustion (°CA)

12
50
Ignition delay (°CA)

10
40
8
30
6

20
4

2 10

0 0
57 99.8 142.5 185.3 228 57 99.8 142.5 185.3 228
Load (Nm) Load (Nm)
Fig. 4. Variations of (a) ignition delay and (b) duration of combustion with engine load.

gine load. For DOC, the increase in load causes increase in duration of
fuel injection and air/fuel mixture formation, and hence the increase in
DOC [26].
Also, Fig. 4(a) and (b) demonstrate that ID is longer for the blended
mode at all engine loads, in comparison with the pure diesel mode;
however the fumigation mode has shorter ID than the blended mode
and could even be shorter than that of the pure diesel mode. On the
average of five loads, the blended mode causes increase in ID (10.1%),
the F + B mode has almost similar ID (only slight rise of 1.4% which
was not significant with T-test at 95% level) while the fumigation mode
has a slight decrease in ID (−1.9%), compared to the pure diesel mode.
The longer ID of the blended mode is due the lower cetane number
and higher latent heat of evaporation of ethanol which cause increase in
ID and decrease in in-cylinder temperature, respectively, compared to
the diesel mode [27,28]. In the fumigation mode, pure diesel which has
higher cetane number than the blended fuel is injected into the mixture
of air and fumigated fuel; in addition, the combustion mixture is also
richer than that in the other modes (Fig. 5) due to the highest fuel
consumption in the fumigation mode. Both factors contribute to shorter
ID in the fumigation mode than in the blended mode. These effects also
are applicable for the F + B mode at only 228 Nm, at which the com-
bustion temperature is high enough resulting in shorter ID. On the other
hand, the longer ID of the fumigation mode at 57 Nm (low load) is due
to lower in-cylinder temperature as a consequence of higher latent heat
of evaporation of ethanol, which dominates over the effect of cetane
number and equivalence ratio. In addition, the too lean mixture of fu-
migated fuel and air (not the overall equivalence ratio as shown in
Fig. 5) plus low combustion temperature at low engine load (57 Nm)
cause increase in ID for the fumigation mode.
For DOC in Fig. 4(b), on the average of five loads, the blended mode
in fuel burned in the diffusion combustion phase (Figs. 2 and 3(b)),
resulting in longer DOC. In addition, the fumigated fuel is not well
atomized (due to low fuel injector pressure of 0.35 MPa) and fuel
mixing with air is poor; therefore, more time is required to burn the
fumigated fuel, resulting in higher DOC in the fumigation mode.
3.2. Engine performance
As shown in Fig. 5, the equivalence ratio (calculated according to
[16]) of all tested fueling modes increases with rise in engine load due
to increase in fuel consumption. On the average of five loads, increases
in equivalence ratio are 9% in the fumigation mode, 7% in the F + B
mode and 2.5% in the blended mode compared to the pure diesel mode.
The higher equivalence ratio of the blended, F + B and fumigation
modes is due to the lower calorific values of biodiesel and ethanol
compared to the pure diesel. The equivalence ratio was the highest with
the fumigation mode for engine loads of 57 Nm to 185.2 Nm because of
incomplete combustion of the fumigated fuel. However, at the engine
load of 228, the equivalence ratio of the fumigated mode was only
slightly less than that of the F + B mode, probably due to improvement
in combustion of the fumigated fuel at the highest engine load. The rise
in equivalence ratio using ethanol fumigation was also reported in
[32,33].
0.7 Diesel
Blend
F+B
0.6 Fumigation
0.5
Equivalence ratio

has shorter DOC (−3.1%), the F + B mode has similar DOC (only slight
rise of 1% which is not significant with T-test at 95% level) and the 0.4
fumigation mode has longer DOC (3.4%), compared to the pure diesel
mode. The reduction in DOC in the blended mode is due to the fol- 0.3
lowing reasons. Firstly, ethanol in the blended fuel causes longer igni-
tion delay and hence increase in heat released in the premixed com- 0.2
bustion phase [29]. Therefore, more fuel is burned during the premixed
combustion phase (Figs. 2 and 3(b)). Secondly, ethanol and biodiesel 0.1
lead to increase in oxygen content of the blended fuel hence decrease
the pyrolysis process and increase the rate of oxidation during com-
0.0
bustion [30]. Jamrozik et al. [31] also found a reduction in DOC and an 57 99.8 142.5 185.3 228
increase in ID with DBE blends.
However, the shorter ID (Fig. 4(a)) of the fumigation mode causes Load (Nm)
decrease in fuel burned in the premixed combustion phase and increase Fig. 5. Variation of equivalence ratio with engine load.

294
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

a) Diesel
b)
40 Diesel
Blend
F+B
Fumigation
350 Blend
F+B 35
Fumigation
300
30
250
BSFC (g/kW.h)

25

BTE (%)
200
20

150 15

100 10

50 5

0 0
57 99.8 142.5 185.3 228 57 99.8 142.5 185.3 228
Load (Nm) Load (Nm)
Fig. 6. Variations of (a) BSFC and (b) BTE with engine load.

F + B mode, with equal BTE (only slight increase of 0.6% which is not
1000 Diesel significant with T-test at 95% level) in the blended mode, in compar-
Blend
F+B ison with the pure diesel mode. The lower calorific value (due to using
Fumigation biodiesel and ethanol) and lower density (due to use of ethanol) lead to
800 increases in BSFC, as compared to the diesel mode [34–36], because
more fuel is required in order to maintain the same output power.
BSCO2 (g/kW.h)

In the blended mode, despite of increase in BSFC (5%), the BTE is

600
400
200
0
57 99.8 142.5 185.3
Load (Nm)
Fig. 7. Variation of BSCO2 with engine load.
According to Fig. 6(a) and (b), the best engine performance, cor-
responding to the lowest BSFC and the highest BTE, occurs at the engine
load of 185.3 Nm for all the tested fueling modes. In contrast, the worst
engine performance takes place at the lowest engine load of 57 Nm for
all the fueling modes due to incomplete combustion, as a consequence
of the low combustion temperature associated with the low fuel air
ratio. On the other hand, at the very high engine load of 228 Nm, the
combustion temperature is high and the fuel/air ratio is also much ri-
cher (Fig. 5), there is not enough time for mixing of fuel and air, re-
sulting in incomplete combustion, decreases in heat release (Fig. 3(b))
and BTE, and increase in BSFC, except for the fumigation mode in BTE
and BSFC. The effect of engine load on BTE (Fig. 6(b)) is similar to that
on HRR (Fig. 3(b)). The BTE increases from the increasing engine load
of 57 Nm to 185.3 Nm and then drops when the engine load is further
increased to 228 Nm for all the fueling modes, except for the fumigation
mode which has almost the same BTE at both engine loads of 185.3 Nm
and 228 Nm.
Fig. 6(a) and (b) also reveal that the blended, F + B and fumigation
modes cause increases in BSFC and drops in BTE (except blended mode
which has similar BTE) compared to the pure diesel mode. On the
average of five loads, the increases in BSFC are 11.2% in the fumigation
mode, 9% in the F + B mode and 5% in the blended mode. The de-
creases in BTE are −5.8% in the fumigation mode and -3.6% in the
similar to diesel mode (0.6%) due to the lower viscosity, improvement
in fuel atomization and increase in oxygen content which improve the
combustion process [15]. However, in the fumigation mode, the fuel
injected into the intake manifold is not well atomized (due to low fuel
injector pressure of 0.35 MPa) resulting in poor fuel mixing with air and
the fumigated fuel cause the highest reduction in in-cylinder tempera-
ture; therefore, incomplete combustion occurs resulting in increase in
BSFC and drop in BTE compared to the those of diesel mode. Şahin et al.
[12] also found that the fumigation mode had higher BSFC and lower
228 BTE than the blended mode, using n-butanol. For the F + B mode, the
results show that it achieves the impact of both blended mode and fu-
migation mode, simultaneously on the BSFC and BTE. Because, the
changes in BSFC (9%) and BTE (−3.6%) of the F + B mode are between
those of the blended and fumigation modes.
3.3. Exhaust gaseous emissions
Variation of BSCO2 with engine load for all the tested fueling modes
is presented in Fig. 7. It can be observed it is similar to that of the BSFC
(Fig. 6(a)) at different engine loads. The lowest BSCO2 is found at 185.3
Nm (which has the lowest BSFC) and the highest BSCO2 is observed at
57 Nm (which has the highest BSFC) for all the fueling modes. Despite
of higher BSFC of the blended, F + B and fumigation modes (as shown
in Fig. 6(a)), based on the average of five loads, the BSCO2 of the
blended mode is lower(−3.3%) than that of the pure diesel mode while
the changes of BSCO2 in the F + B mode (−0.5%) and in the fumiga-
tion mode (0.6%), compared to the pure diesel mode, are statistically
insignificant at the 95% confidence level. The reduction in CO2 emis-
sion in the blended mode was reported in the literature [15,29,37,38].
The reduction in CO2 emission in the blended mode and similar CO2
emission in the F + B and fumigation modes is due to the lower carbon-
to-hydrogen ratio and higher oxygen content of biodiesel and ethanol in
comparison with the diesel mode [38–40]. However, the blended mode
has lower BSFC compared to the F + B and fumigation modes, resulting
of lower CO2 formation and hence lower BSCO2. In addition, the lowest
BSCO2 of the blended mode is due to improvement in BTE [11] in
comparison with the other fueling modes.

295
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

hydrocarbons at higher in-cylinder temperatures [42] resulting in lower

Diesel
Blend BSHC emissions in comparison with the diesel mode. The reduction in
14
F+B BSHC is 14.7% in the blended mode at the engine load of 228 Nm. The
Fumigation
rise in BSHC at lower and even medium engine loads and a drop at
12
higher load using DBE blend was observed in other studies [29,43]. For
the fumigation mode, some factors have impact on huge increase in
10
BSCO (g/kW.h)

BSHC emission. Firstly, the large droplet size of the fumigated fuel (due
to low injector pressure) causes incomplete combustion. Secondly, the
8
mixture of fumigated fuel and air causes decrease in combustion tem-
6
perature (production of quenching flame layers), especially at lower
loads resulting in incomplete combustion. Thirdly, fumigation mode
4
has the highest equivalence ratio (Fig. 5) and fuel consumption
(Fig. 6(a)) resulting more fuel in combustion. Finally, some fraction of
2 fumigated fuel could easily escape into the low temperature quenching
regions close to the combustion chamber walls and crevices (like the
0 region between the piston crown and the cylinder wall) especially
57 99.8 142.5 185.3 228 during the compression stroke and expansion stroke, which becomes
one of the primary sources of unburned hydrocarbon emission [44,45].
Load (Nm) The higher HC emission in the fumigation mode compared to that in the
Fig. 8. Variation of BSCO with engine load. blended mode was also reported in the literature [11,12].
It can be seen from Fig. 10 that the BSNOX approximately decreases
with rise in engine load which is in line with results reported in other
Fig. 8 depicts that BSCO emission decreases with rise in engine load
investigations using diesel and DBE blend [15,26] and using fumigation
for all the tested fueling modes due to increase in combustion tem-
mode [11]. Fig. 10 also reveals that the blended, F + B and fumigation
perature and more complete combustion. Similar results were also re-
modes have lower NOX at all engine loads compared to the pure diesel
ported for diesel and blended fuels [29] and for operation in the fu-
mode. According to the average of five loads the reductions in NOX
migation mode [32,41] modes using ethanol. It can be seen from Fig. 8
emissions are −8.6% in fumigation mode, −5.6% in F + B mode and
that the blended, F + B and fumigation modes, in comparison with the
−3.6% in blended mode compared to that of the pure diesel mode. The
diesel mode, cause increases in BSCO emission at all tested engine
higher latent heat of evaporation and lower heating value of biodiesel
loads, except at 185.3 and 228 Nm for the blended fuel. In the blended
and ethanol are the more dominating factors, compared to the other
mode, at low load, the lower combustion temperature associated with
parameters (like lower cetane number and higher oxygen content of
the cooling effect of ethanol is dominating, which suppresses the CO
biodiesel and ethanol which increase the combustion temperature),
oxidation process, resulting in rise in BSCO in comparison with the
which cause reduction in NOX for all the tested fueling modes. The
diesel mode [6,38]. The increment of BSCO is 39.3% using blended
lowest BSNOX of the fumigation mode is due to the lower combustion
mode at engine load of 57 Nm. However, with increase in engine load,
temperature as a consequence of incomplete combustion compared to
the impact of combustion temperature becomes weaker while the im-
the other fueling modes. In the fumigation mode, the fumigated fuel is
pact of oxygen content of the fuel becomes more dominating, hence the
injected into the cylinder via intake manifold which absorbs the heat
BSCO of the blended mode is lower than that of diesel mode at the
inside the combustion chamber, causing a lower temperature environ-
higher loads of 185.3 and 228 Nm (Fig. 8). The reduction in BSCO is
ment for combustion; therefore incomplete combustion takes place re-
1.3% at the engine load of 228 Nm in the blended mode.
sulting in higher CO, HC and lower NOX emissions. Higher effect of the
The BSCO is the highest in the fumigation mode. In the fumigation
fumigation mode on reduction in NOX emission compared to the
mode, the effect of incomplete combustion, lower combustion tem-
blended mode was reported in the literature [11,12].
perature and higher BSFC lead to higher CO emission at all the tested
It is noticeable that, despite of reduction in BSNOX when using
loads. The higher CO emission of fumigation mode compared to
blended, F + B or fumigation modes of operation, BSNO2 emissions are
blended mode was also found in [11]. On the average of five loads, the
F + B mode causes 87.7% rise in BSCO which is between those of the
blended mode (17%) and the fumigation mode (152.9%), compared to
12 Diesel
the diesel. Blend
Fig. 9 shows the results for BSHC which are similar to those for F+B
Fumigation
BSCO (Fig. 8). BSHC emission decreases with rise in engine load and the 10
use of blended, F + B and fumigation modes cause higher BSHC emis-
sion at all tested engine loads (except at 185.3 and 228 Nm for the
8
BSHC (g/kW.h)

blended fuel).
According to Fig. 9, on the average of five loads, the blended mode
(4.6%), F + B mode (86.5%) and fumigation mode (158.2%) have 6
higher BSHC emission compared to that of the pure diesel mode. It is
noticeable that, similar to BSCO emission, the BSHC emission of the
4
F + B mode is between that of the blended mode and the fumigation
mode.
For the blended mode, the rise in BSHC at the low engine load 2
(similar to BSCO) is due to incomplete combustion as a consequence of
the higher latent heat of evaporation of ethanol compared to the diesel,
0
which lead to 22.7% increase in BSHC at the engine load of 57 Nm. But, 57 99.8 142.5 185.3 228
the higher oxygen content of ethanol and biodiesel in the blended fuel
is the more dominating factor at higher engine loads which causes more Load (Nm)
complete combustion and increase the oxidation of unburned Fig. 9. Variation of BSHC with engine load.

296
M.A. Ghadikolaei et al.
5 Diesel
Blend
F+B
Fumigation
4 mation [42]. In addition, incomplete combustion due to lack of enough
BSNOX (g/kW.h)
3 Fig. 12 also illustrates that the BSPM emissions of the blended and
2
1
0 [9,32,48,49]. The increase in PM mass in low load (idle) and decrease
57 99.8 142.5 185.3 228
Load (Nm)
Fig. 10. Variation of BSNOX with engine load.
increased by 187.2% in the fumigation mode, 111.5% in the F + B
mode and 3.3% in the blended mode, compared to that of the pure
diesel mode, based on the average of five engine loads (Fig. 11). This
finding is in line with [11], which found the BSNO2 emissions of the
blended and fumigation modes were higher than that of the baseline
fuel and the impact of fumigation mode on increase in BSNO2 was
higher than that of blended mode.
The NO formed in the flame zone can be rapidly converted to NO2
through the reaction (9) [44] when the fluid is cooled:
NO + HO2 → NO2 + OH
The cooler regions with the increases of engine load are reduced
resulting in less conversion of NO into NO2 [46], therefore the NO2
emission drops with increase in engine load (Fig. 11). The increases in
BSNO2 emissions when using the blended, F + B and fumigation modes
can be attributed to two reasons. Firstly, in hydrocarbon decomposition
processes, ethanol is a well-known chain-carrying agent which can
react with OH radicals to reform the peroxyl radicals (HO2); thus, the
ethanol is a source of HO2 [47] which increases the formation on NO2
through Eq. (9). Secondly, the lower combustion temperature due to
ethanol can enhance the conversion of NO into NO2 through Eq. (9).
Therefore, the BSNO2 emissions of the blended, F + B and fumigation
modes are higher than that of the pure diesel mode. While, the highest
BSNO2 achieved by the fumigation mode is due to the lowest com-
bustion temperature.
3.4. PM emissions
In regard to the investigation of PM emissions, the effects on BSPM
mass, total number concentration (TNC) and geometric mean diameter
(GMD) are analyzed. Fig. 12 shows that the BSPM emissions of all the
tested modes decrease slightly from the engine load of 57 Nm to
99.8 Nm load and then increase up to the highest engine load of 228
Nm. This trend was reported also in the literature using DBE and diesel
modes [15,18] and alcohol fumigation (ethanol and methanol) [48,49].
At the lowest engine load, the combustion is incomplete due to the low
combustion temperature, therefore, BSPM is higher than that at the
engine load of 99.8 Nm. On the other hand, the increase in BSPM mass
at high engine load can be attributed to the following factors. At high
engine load, the amount of fuel burned is increased and the available
time for soot oxidation after the end of the diffusion combustion is low
[50] resulting in rise in particulate formation as well as reduction in
time for soot oxidation, and thus increase in BSPM. Moreover, the
higher F/A ratio results in lower oxygen concentration availability for
Fuel 235 (2019) 288–300
soot oxidation at high loads. Also, the ignition delay becomes shorter
with rise in engine load; therefore, more fuel is burned during the
diffusion combustion period, resulting in increase in particulate for-
mixing time for fuel and air causes an increase in PM mass at high
engine load.
F + B modes are lower than that of the pure diesel mode at all tested
engine loads. However, the fumigation mode has lower BSPM only at
higher engine loads (185.3 and 228 Nm). On the average of five loads,
the reductions in BSPM are −47.9% in the blended mode and −28.4%
in the F + B mode; while the fumigation mode has a similar BSPM (only
slight increase of 1.2% which is not significant with T-test at 95% level)
compared to that of the diesel mode. The reductions in PM mass can
also be found in the literature when using DBE in the blended mode
[7,18,51] and using ethanol and methanol in the fumigation mode
in PM mass at higher engine load was also reported in [49] using fu-
migated ethanol compared to the diesel.
There are several factors leading to the reduction in BSPM when
using biofuels. Firstly, the carbon to hydrogen mass ratios of the bio-
fuels are lower than that of diesel fuel resulting in decrease in PM mass
[42]. Secondly, the lower aromatic and sulfur contents of these fuels
lead to decrease in PM emission [42]. Thirdly, higher oxygen content of
these fuels enhances the soot oxidation through the hydroxyl radical (%
OH) formation which leads to lower soot formation [7].
However, in the fumigation mode, several factors lead to increase in
PM emissions at low and medium load. Firstly, the fumigated fuel is
uniformly mixed with air inside the engine cylinder; while, at low and
medium engine loads, the equivalence ratio of the uniformly mixed air/
fuel mixture is very low, and the combustion temperature is also low,
(9) leading to very poor combustion of the fumigated fuel. Part of the fuel
might enter into the engine exhaust and condense during the cooling
down process in the PM sampling process, leading to an increase in the
PM emission. The increase in unburn fuel emission at low and medium
engine loads could also be observed in the high HC emissions in the
fumigation mode of operation (Fig. 9). Secondly, part of the fumigated
fuel could easily enter into the low temperature quenching regions close
to the combustion chamber walls and crevices which can produce
higher PM mass (especially during the low and medium loads). Thirdly,
the injection pressure for the fumigation mode was low (0.35 MPa),
therefore the large fuel droplets cause formation of more particles and
larger particles (as shown in TNC and GMD at Figs. 13 and 14, re-
spectively) and hence increase in particle mass. Fourthly, the shorter ID
3.0
Diesel
Blend
F+B
2.5
Fumigation
2.0
BSNO2 (g/kW.h)
1.5
1.0
0.5
0.0
57 99.8 142.5 185.3 228
Load (Nm)
Fig. 11. Variation of BSNO2 with engine load.
297
M.A. Ghadikolaei et al. Fuel 235 (2019) 288–300

500 and longer DOC in the fumigation mode (Figs. 2 and 4(a) and (b)) lead
Diesel
Blend
to increase in the diffusion combustion period. Finally, fumigation
F+B mode has higher BSFC and overall equivalence ratio which cause more
400 Fumigation
fuel available for PM formation. However, at higher engine loads
(185.3 and 228 Nm), the above reasons have no effect due to more
complete combustion, as a consequence of higher in-cylinder tem-
BSPM (mg/kW.h)

300 perature and higher equivalence ratio; therefore the BSPM of the fu-
migation mode are lower than those of diesel mode. For the F + B
mode, the results show that it has a reduction of −28.4% in PM mass,
200 which is located between those of the other two modes of operation.
The reduction in PM mass can be found in the literature when using
DBE in the blended mode [7,18,51] and ethanol and methanol in the
100 fumigation mode [9,32,48,49]. However, in the fumigation mode in the
present study, on the average of five loads, the fumigation mode has a
similar PM mass compared to the diesel mode which can be due to use
of different types and percentages of fumigated fuel and different op-
0
57 99.8 142.5 185.3 228 erating conditions (engine loads and speeds); while, the reduction in
PM mass in this study was found only at higher engine loads.
Load (Nm) In regard to TNC and GMD of the particles, Figs. 13 and 14 reveal
Fig. 12. Variation of BSPM with engine load. that both of them increase with rise in engine load for all the tested
fueling modes which are in line with those reported in the literature
when using DBE in the blended mode [15] and ethanol and methanol in
the fumigation mode [20,48,52,53]. The same factors which lead to
increase in PM mass also lead to increase in TNC. For GMD, at high
Total Number Concentration x E+7 (#/cm³)

Diesel
Blend engine loads, the huge number of particles causes increase in particles
10 F+B
coagulation rate and hence particles with larger GMD are formed [54].
Fumigation
Figs. 13 and 14 also illustrate that the blended and F + B modes
have lower TNC (except almost similar TNC at lower and medium loads
8
for F + B) and GMD at all the tested loads compared to those of the
pure diesel mode. However, the fumigation mode has lower TNC and
6 GMD only at higher engine loads (185.3 and 228 Nm). Zhang et al. [53]
also found that the GMD of fumigation operation with methanol was
higher at low and medium engine loads and lower at high engine loads
4 compared to the diesel mode. Similarly, the increase in GMD at lower
engine load and decrease in GMD at higher engine load using fumigated
ethanol was reported in another work [32]. On the average of five
2
loads, the blended mode has lower TNC (−18.5%) and the F + B mode
has almost similar TNC (only −1.2%) while the fumigation has higher
0 TNC (27.8%) than diesel mode. For GMD, the blended mode (−9.3%)
57 99.8 142.5 185.3 228 and the F + B mode (-4.8%) cause reductions while the fumigation
mode has almost equal GMD (only 1.5% which was not significant with
Load (Nm) T-test at 95% level), compared to those of the pure diesel mode. Lower
Fig. 13. Variation of total particle number concentration with engine load. aromatic and sulfur contents and carbon to hydrogen mass ratio [42],
higher oxygen content (for enhance in soot oxidation) [7] of the
blended, F + B and fumigation (only at higher engine loads) fuels cause
100 reductions in TNC and GMD compared to those of the pure diesel mode.
Diesel
Blend The reductions in TNC and GMD can also be found in the literature
F+B when using DBE [18,51] and the fumigation mode (ethanol and me-
Geometric Mean Diameter (nm)

Fumigation
80 thanol) [20,32,48,53,55]. On the other hand, an increase in TNC was
reported in [49] using fumigated ethanol at all the tested loads com-
pared to the diesel. In the present study, the reductions in TNC and
60 GMD were found only at higher engine loads. For fumigation mode at
low and medium loads, the same reasons leading to the increase in PM
mass also lead to the increase in TNC (Fig. 13) and GMD (Fig. 14)
40 compared to the diesel mode. However, at higher engine loads, the
combustion is complete as a consequence of higher in-cylinder tem-
perature in fumigation mode resulting in lower productions of TNC and
20 GMD in comparison with those of the diesel mode.

3.5. Summarized effects

0
57 99.8 142.5 185.3 228 A comparative summary of the influences of the different fueling
modes on the combustion, performance and emissions based on average
Load (Nm) of five loads are listed in Table 4. It can be seen that the blended mode
Fig. 14. Variation of geometric mean diameter with engine load. and fumigation mode have the best and worst effects on the engine
performance and emissions, respectively, except that NOx emission is

298
M.A. Ghadikolaei et al.
Table 4
Comparative summary of tested parameters for all fueling modes based on average of five loads.
Parameter
Peak in-cylinder pressure
Peak HRR
ID
DOC
Equivalence ratio
BSFC
BTE
BSCO2
BSCO
BSHC
BSNOX
BSNO2
BSPM
TNC
GMD
lower in the fumigation mode than in the blended mode. In addition,
the values for all the parameter for F + B mode are located between
those of the blended and fumigation modes. Overall, the application of
the biofuels, in either the blended, fumigation or F + B mode, in
comparison with the diesel mode, could lead to reduction in NOx and
PM (except fumigation mode) emissions, but increase in CO and HC
emissions as well as fuel consumption.
4. Conclusions
The present experimental study was conducted to investigate and
compare the effects of using diesel, biodiesel and ethanol in the blended
and fumigation modes on the combustion, performance and emissions
of a diesel engine. In addition, this study introduced a combined fu-
migation with blended mode (F + B). The experiments were conducted
at a constant fuel ratio of 80% diesel, 5% biodiesel and 15% ethanol in
the blended, F + B and fumigation modes to provide the same fuel
composition for comparing the effects of the three fueling modes.
In regard to engine combustion, it was found that, the blended mode
caused increases in peak heat release rate (20.4%) and ignition delay
(10.1%), and decrease in duration of combustion (−3.1%), and the
same peak in-cylinder pressure (−0.2%) in comparison with those
obtained from the pure diesel mode. Also, the fumigation mode had
higher peak HRR (2.2%) and DOC (3.4%), lower ID (−1.9%), and si-
milar peak in-cylinder pressure (1.3%) compared to those of the pure
diesel mode. In addition, the F + B mode had higher peak HRR (6.3%),
and similar peak in-cylinder pressure (0.9%), ID (1.4%) and DOC (1%)
compared to those of the pure diesel mode.
For engine performance, the fumigation mode had the worst engine
performance because of the highest BSFC and the lowest BTE among all
the tested fueling modes. The increase in BSFC is 11.2% in the fumi-
gation mode, 9% in the F + B mode and 5% in the blended mode
compared to the pure diesel mode. And the reduction in BTE is -5.8% in
the fumigation mode and -3.6% in the F + B mode while the blended
mode had similar BTE (only 0.6% rise) with the pure diesel mode.
In respect of emissions, the fumigation mode caused increases in CO
(152.9%), HC (158.2%), NO2 (187.2%) and TNC (27.8%), and reduc-
tion only in NOX (−8.6%), and almost equal CO2 (0.6%), PM (1.2%)
and GMD (1.5%), compared to the pure diesel mode. However, the
blended mode had lower CO2 (−3.3%), NOX (−3.6%), PM (−47.9%),
TNC (−18.5%), GMD (−9.3%) and higher CO (17%), HC (4.6%) and
NO2 (3.3%) compared to the pure diesel mode.
The results illustrated that the F + B mode had the emission values
almost in the middle of the blended and fumigation modes. In detail,
the F + B mode had higher CO (87.7%), HC (86.5%) and NO2
(111.5%), and lower NOX (−5.6%), PM (−28.4%), GMD (−4.8%), and
similar TNC (−1.2%), CO2 (−0.5%), in comparison with those of the
Fuel 235 (2019) 288–300
Order (Highest to lowest)
Fumigation (1.3%) ≈ F + B (0.9%) ≈ Diesel ≈ Blend (−0.2%)
Blend (20.4%) > F + B (6.3%) > Fumigation (2.2%) > Diesel
Blend (10.1%) > F + B (1.4%) ≈ Diesel > Fumigation (−1.9%)
Fumigation (3.4%) > F + B (1%) ≈ Diesel > Blend (−3.1%)
Fumigation (9%) > F + B (7%) > Blend (2.5%) > Diesel
Fumigation (11.2%) > F + B (9%) > Blend (5%) > Diesel
Blend (0.6%) ≈ Diesel > F + B (−3.6%) > Fumigation (−5.8%)
Fumigation (0.6%) ≈ Diesel ≈ F + B (−0.5%) > Blend (−3.3%)
Fumigation (152.9%) > F + B (87.7%) > Blend (17%) > Diesel
Fumigation (158.2%) > F + B (86.5%) > Blend (4.6%) > Diesel
Diesel > Blend (−3.6%) > F + B (−5.6%) > Fumigation (−8.6%)
Fumigation (187.2%) > F + B (111.5%) > Blend (3.3%) > Diesel
Fumigation (1.2%) ≈ Diesel > F + B (−28.4%) > Blend (−47.9%)
Fumigation (27.8%) > Diesel ≈ F + B (−1.2%) > Blend (−18.5%)
Fumigation (1.5%) ≈ Diesel > F + B (−4.8%) > Blend (−9.3%)
pure diesel mode.
To sum up, it can be inferred from the results that the blended mode
shows the highest engine performance and lowest emissions (except for
BSNOX) among the blended, F + B and fumigation modes. In contrast,
the fumigation mode is the worst fueling mode in terms of engine
performance and emissions (except for BSNOX). However, the F + B
mode has engine performance and emissions characteristics between
those of the blended mode and the fumigation modes.
Acknowledgement
The authors would like to thank The Hong Kong Polytechnic
University for the financial support (RUAT).
References
[1] Murad Algayyim SJ, Wandel AP, Yusaf T, Hamawand I. Production and application
of ABE as a biofuel. Renew Sustain Energy Rev 2018;82(1):1195–214.
[2] Mofijur M, Rasul MG, Hyde J, Azad AK, Mamat R, Bhuiya MMK. Role of biofuel and
their binary (diesel-biodiesel) and ternary (ethanol-biodiesel-diesel) blends on in-
ternal combustion engines emission reduction. Renew Sustain Energy Rev
2016;53:265–78.
[3] Shahir SA, Masjuki HH, Kalam MA, Imran A, Rizwanul Fattah IM, Sanjid A.
Feasibility of diesel-biodiesel-ethanol/bioethanol blend as existing CI engine fuel:
an assessment of properties, material compatibility, safety and combustion. Renew
Sustain Energy Rev 2014;32:379–95.
[4] Zöldy M. Ethanol-biodiesel-diesel blends as a diesel extender option compression
ignition engines. Transport 2011;26:303–9.
[5] Barabas I, Todoruţ IA. Utilization of biodiesel-diesel-ethanol blends in CI engine. In:
Montero G, Stoytcheva M, editors. Biodiesel-Quality, Emissions and by-Products.
InTech; 2011. p. 215–34.
[6] Hulwan DB, Joshi SV. Performance, emission and combustion characteristic of a
multicylinder DI diesel engine running on diesel-ethanol-biodiesel blends of high
ethanol content. Appl Energy 2011;88(12):5042–55.
[7] Shahir SA, Masjuki HH, Kalam MA, Imran A, Ashraful AM. Performance and
emission assessment of diesel-biodiesel-ethanol/bioethanol blend as a fuel in diesel
engines: a review. Renew Sustain Energy Rev 2015;48:62–78.
[8] Imran A, Varman M, Masjuki HH, Kalam MA. Review on alcohol fumigation on
diesel engine: A viable alternative dual fuel technology for satisfactory engine
performance and reduction of environment concerning emission. Renew Sustain
Energy Rev 2013;26:739–51.
[9] Ghadikolaei MA. Effect of alcohol blend and fumigation on regulated and un-
regulated emissions of IC engines – a review. Renew Sustain Energy Rev
2016;57:1440–95.
[10] Abu-Qudais M, Haddad O, Qudaisat M. The effect of alcohol fumigation on diesel
engine performance and emissions. Energy Convers Manage 2000;41:389–99.
[11] Cheng CH, Cheung CS. Chan TL, Lee SC, Yao CD, Tsang KS. Comparison of emis-
sions of a direct injection diesel engine operating on biodiesel with emulsified and
fumigated methanol. Fuel 2008; 87:1870-1879.
[12] Şahin Z, Durgun O, Aksu ON. Experimental investigation of n-butanol/diesel fuel
blends and n-butanol fumigation – Evaluation of engine performance, exhaust
emissions, heat release and flammability analysis. Energy Convers Manage
2015;103:778–89.
[13] Mariasiu F, Burnete NV, Moldovanu D, Varga BO, Iclodean C, Kocsis L. Effects of
bioethanol ultrasonic generated aerosols application on diesel engine performances.
Therm Sci 2015;19(6):1931–41.
[14] Man XJ, Cheung CS, Ning Z, Yung KF. Effect of waste cooking oil biodiesel on the
299
M.A. Ghadikolaei et al.
properties of particulate from a DI diesel engine. Aerosol Sci Technol
2015;49(4):199–209.
[15] Tse H, Leung CW, Cheung CS. Performances, emissions and soot properties from a
diesel-biodiesel ethanol blend fuelled engine. Adv Automob Eng 2016;S1:005.
https://doi.org/10.4172/2167-7670.S1-005.
[16] Bresenham D, Reisel J, Neusen K. Spindt air-fuel ratio method generalization for
oxygenated fuels. SAE Technical Paper 1998; Paper No: 982054. doi: https://doi.
org/10.4271/982054.
[17] Moffat RJ. Describing the uncertainties in experimental results. Exp Therm Fluid Sci
1988;1:3–17.
[18] Tse H, Leung CW, Cheung CS. Investigation on the combustion characteristics and
particulate emissions from a diesel engine fueled with diesel-biodiesel-ethanol
blends. Energy 2015;83:343–50.
[19] Qi DH, Chen H, Geng LM, Bian YZH, Ren XCH. Performance and combustion
characteristics of biodiesel-diesel-methanol blend fuelled engine. Appl Energy
2010;87:1679–86.
[20] Zhang ZH, Cheung CS, Yao CD. Influence of fumigation methanol on the combus-
tion and particulate emissions of a diesel engine. Fuel 2013;111:442–8.
[21] Zhu L, Cheung CS, Zhang WG, Huang Z. Combustion, performance and emission
characteristics of a DI diesel engine fueled with ethanol-biodiesel blends. Fuel
2011;90:1743–50.
[22] Wang LJ, Song RZ, Zou HB, Liu SH. Zhou LB. Study on combustion characteristics of
a methanol–diesel dual-fuel compression ignition engine. Proceed Institut Mech Eng
Part DJ Automob Eng 2008;222:619–27.
[23] Zhu L, Cheung CS, Zhang WG, Huang Z. Emissions characteristics of a diesel engine
operating on biodiesel and biodiesel blended with ethanol and methanol. Sci Total
Environ 2010;408(4):914–21.
[24] Rajesh Kumar B, Saravanan S, Rana D, Nagendran A. A comparative analysis on
combustion and emissions of some next generation higher-alcohol/diesel blends in
a direct-injection diesel engine. Energy Convers Manage 2016;119:246–56.
[25] Park SH, Youn IM, Lee CS. Influence of ethanol blends on the combustion perfor-
mance and exhaust emission characteristics of a four-cylinder diesel engine at
various engine loads and injection timings. Fuel 2011;90:1748–55.
[26] Hansdah D, Murugan S. Bioethanol fumigation in a DI diesel engine. Fuel
2014;130:324–33.
[27] Zheng Z, Li C, Liu H, Zhang Y, Zhong X, Yao M. Experimental study on diesel
conventional and low temperature combustion by fueling four isomers of butanol.
Fuel 2015;141:109–19.
[28] Shi X, Pang X, Mu Y, He H, Shuai S, Wang J, et al. Emission reduction potential of
using ethanol-biodiesel-diesel fuel blend on a heavy-duty diesel engine. Atmos
Environ 2006;40:2567–74.
[29] Oliveira AD, Morais AMD, Valente OS, Sodré JR. Combustion characteristics, per-
formance and emissions from a diesel power generator fuelled by B7-ethanol
blends. Fuel Process Technol 2015;139:67–72.
[30] Donahue R, Foster DE. Effects of oxygen enhancement on the emissions from a DI
diesel via. Manipulation of fuels and combustion chamber gas composition. SAE
Technical Paper 2000; No. 2000-01-0512.
[31] Jamrozik A, Tutak W, Pyrc M, Sobiepanski M. Effect of diesel-biodiesel-ethanol
blend on combustion, performance, and emissions characteristics on a direct in-
jection diesel engine. Therm Sci 2017;21:591–604.
[32] López AF, Cadrazco M, Agudelo AF, Corredor LA, Vélez JA, Agudelo JR. Impact of
n-butanol and hydrous ethanol fumigation on the performance and pollutant
emissions of an automotive diesel engine. Fuel 2015;153:483–91.
[33] Britto RF, Martins CA. Experimental analysis of a diesel engine operating in diesel-
ethanol dual-fuel mode. Fuel 2014;134:140–50.
[34] Pidol L, Lecointe B, Starck L, Jeuland N. Ethanol-biodiesel-Diesel fuel blends:
Performances and emissions in conventional diesel and advanced low temperature
300
Fuel 235 (2019) 288–300
combustions. Fuel 2012;93:329–38.
[35] Atmanli A. Comparative analyses of diesel-waste oil biodiesel and propanol, n-bu-
tanol or 1-pentanol blends in a diesel engine. Fuel 2016;176:209–15.
[36] Rajesh Kumar B, Saravanan S. Use of higher alcohol biofuels in diesel engines: A
review. Renew Sustain Energy Rev 2016;60:84–115.
[37] Noorollahi Y, Azadbakht M, Ghobadian B. The effect of different diesterol (diesel-
biodiesel-ethanol) blends on small air-cooled diesel engine performance and its
exhaust gases. Energy 2018;142:196–200.
[38] He BQ, Shuai SJ, Wang JX, He H. The effect of ethanol blended diesel fuels on
emissions from a diesel engine. Atmos Environ 2003;37:4965–71.
[39] Ying W, Longbao Z, Hewu W. Diesel emission improvements by the use of oxyge-
nated DMW/diesel blend fuels. Atmos Environ 2006;40:2313–20.
[40] Paul G, Datta A, Mandal B. Numerical investigation of the performance and emis-
sion characteristics of a CI engine using diesel and its blends with ethanol and
Jatropha biodiesel. Int J Curr Eng Technol 2014;3:5–9.
[41] Surawski NC, Ristovski ZD, Brown RJ, Situ R. Gaseous and particle emissions from
an ethanol fumigated compression ignition engine. Energy Convers Manage
2012;54(1):145–51.
[42] Wei L, Cheung CS, Huang Z. Effect of n-pentanol addition on the combustion,
performance and emission characteristics of a direct-injection diesel engine. Energy
2014;70:172–80.
[43] Yilmaz N, Vigil FM, Donaldson AB, Darabseh T. Investigation of CI engine emissions
in biodiesel-ethanol-diesel blends as a function of ethanol concentration. Fuel
2014;115:790–3.
[44] Heywood JB. Internal combustion engine fundamentals. 2nd ed. New York:
McGraw-Hill; 2003.
[45] Song R, Liu J, Wang L, Liu S. Performance and emissions of a diesel engine fuelled
with methanol. Energy Fuels 2008;22(6):3883–8.
[46] Wei L, Cheng R, Mao H, Geng P, Zhang Y, You K. Combustion process and NOX
emissions of a marine auxiliary diesel engine fuelled with waste cooking oil bio-
diesel blends. Energy 2018;144:73–80.
[47] Chu PK, Lu XP. Low temperature plasma technology: methods and applications.
Florida: CRC Press; 2013.
[48] Zhang ZH, Tsang KS, Cheung CS, Chan TL, Yao CD. Effect of fumigation methanol
and ethanol on the gaseous and particulate emissions of a direct-injection diesel
engine. Atmos Environ 2011;45(11):2001–8.
[49] Surawski NC, Miljevic B, Roberts BA, Modini RL, Situ R, Brown RJ, et al. Particle
emissions, volatility, and toxicity from an ethanol fumigated compression ignition
engine. Environ Sci Technol 2010;44(1):229–35.
[50] Tsolakis A. Effects on particle size distribution from the diesel engine operating on
RME-biodiesel with EGR. Energy Fuels 2006;20:1418–24.
[51] Muralidharan M, Subramanian M, Kanal PC, Malhotra RK. Evaluation of a novel
biofuel blend using diesel-biodiesel-ethanol on light commercial vehicle. SAE Paper
2011-28-0015. doi: 10.4271/2011-28-0015.
[52] Cheng CH, Cheung CS, Chan TL, Lee SC, Yao CD. Experimental investigation on the
performance, gaseous and particulate emissions of a methanol fumigated diesel
engine. Sci Total Environ 2008;389:115–24.
[53] Zhang ZH, Cheung CS, Chan TL, Yao CD. Emission reduction from diesel engine
using fumigation methanol and diesel oxidation catalyst. Sci Total Environ
2009;407:4497–505.
[54] Di Y, Cheung CS, Huang Z. Experimental investigation on regulated and un-
regulated 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.
[55] Zhang ZH, Cheung CS, Chan TL, Yao CD. Experimental study on particulate emis-
sions of a methanol fumigated diesel engine equipped with diesel oxidation catalyst.
Aerosol Sci Technol 2011;45:262–71.