Fuel 107 (2013) 409–418

Contents lists available at SciVerse ScienceDirect

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

Combustion and emission characteristics of n-butanol/diesel fuel blend

in a turbo-charged compression ignition engine
Lennox Siwale a,⇑, Lukács Kristóf b, Torok Adam b, Akos Bereczky b, Makame Mbarawa c,
Antal Penninger b, Andrei Kolesnikov a
a
Department of Mechanical Engineering, Tshwane University of Technology, Private Bag X680, Pretoria 0001, South Africa
b
Department of Energy Engineering, Budapest University of Technology and Economics, H-1111 Budapest, Bertalan Lajos u. 4–6, D208, Hungary
c
Ministry of Communication, Science and Technology, Jamhuri Street, P.O. Box 2645, Dar es Salaam, United Republic of Tanzania

h i g h l i g h t s

" Regulated emissions were compared with other similar study by other authors using biodiesel blending with diesel fuel.
" There was great improvement in the reduction of the emissions using a small shared volume of n-butanol with diesel fuel.
" Premixed combustion phase is distinguishable and amplified suggesting an effective air-to-fuel mixing process.
" We recommended the use of n-butanol/diesel blend in a turbo-charged engine without any engine modifications.

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

Article history: Burning of 5%, 10%, and 20% shared volume of n-butanol (B05, B10, and B20, where B05 represents 5%
Received 14 August 2012 shared volume of n-butanol with 95% diesel fuel) with diesel fuel (DF) in a high load, light duty,
Received in revised form 22 October 2012 turbo-charged diesel engine is reported. The aim was to compare the effects of the blends on the engine
Accepted 27 November 2012
combustion characteristics and regulated emissions namely nitrogen oxides (NOx), unburned hydrocar-
Available online 19 December 2012
bon (UHC), carbon monoxide (CO) and soot results from this study with a similar past study: (30% rape-
seed oil methyl esters (RMEs) shared volume with similar diesel fuel admixed to 5% or 7.5% bioethanol).
Keywords:
Using n-butanol shared volume, B05, B10 and B20 significantly improved the reduction of regulated
Compression ignition
Diesel
emissions compared to the other study. In this study the reduction or increase percentage relative to
Emission reduction DF was as follows at 75% load at 1500 rpm for B5, B10, B20 mixtures: soot reduction was 55.5%, 77.8%,
Butanol (-n) and 85.1% respectively; CO reduction was 35.7%, 57.1% and 71.4%; NOx increase was 10.3%, 32.3% and
Global climate change 54.4% ; UHC increase, 21.4%, 71.4%, 214% respectively. The premixed phase combustion was amplified
and distinguishable with increase of shared volume of n-butanol in DF. The combustion cycles of the
blends were more stable than the cycles of DF.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction such as ethers, esters, and bioalcohols, have been added to diesel
fuel [6]. In gasoline engines biobutanol has been studied and dis-
Environmental degradation of petroleum products and their cussed as potential fuel [7–10]. Some studies exist where vegetable
non-renewable nature has led to a world-wide search for renew- oils were blended with ethanol and n-butanol to test diesel engine
able and greener alternatives in internal combustion. Such alterna- performance. Butanol/diesel was used in a light-duty turbo diesel
tives which include biofuels [1] have in general the advantage of vehicle [4,11]. Different blend ratios and amounts of regulated
reducing most of the regulated emissions such as nitrogen oxides emissions have been reported in literature.
(NOx), unburned hydrocarbons (UHC), soot, carbon monoxide The nature of the fuel [12] has a significant influence on the com-
[2–4] in reciprocating engines. It has been shown [5] that biofuels bustion and emissions generated in internal combustion engines.
and several cell electric vehicles are the best options in reducing To minimize concentration of emissions in exhaust gases, bioalco-
greenhouse gases by 80% or more below the 1990 levels in the hols could be used. Bioalcohols are oxygenated fuels that form part
transportation sector. Biofuels, which include oxygenated additives of the biofuel family. Oxygenated fuels have drawn the attention of
many researchers due to the ability to drastically reduce particulate
⇑ Corresponding author. Tel.: +27 764422347/12 382 5164; fax: +27 12 382 5602. matter (PM) emission without altering the emission levels of other
E-mail address: [email protected] (L. Siwale).
regulated emissions. Most studies have focused the use of ethanol

0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2012.11.083
410 L. Siwale et al. / Fuel 107 (2013) 409–418

Nomenclature

Bx x% volume of n-butanol of blend CA crank angle

BD biodiesel (RME) LHV lower heating value
BMEP brake mean effective pressure ID ignition delay
BSFC brake specific fuel consumption PM particulate matter
DF diesel fuel
EGT exhaust gases temperature Greek letters
TDC top dead center k excess air ratio
BTDC before top dead center h crank angle
ATDC after top dead center
UHC unburned hydrocarbon

as fuel in reciprocating engines. Fewer studies have reported the phates and calcium sulfonates are added in lubricant oil to
use of n-butanol as fuel, although n-butanol posses some better fuel neutralize these acids and improve lubrication. Shorter intervals
properties than ethanol [12]. of lubricant oil change reduce corrosive wear significantly. All alco-
Ethanol, which is highly miscible with gasoline, is hygroscopic hols are highly corrosive, which is a function of water content; but
or less miscible when mixed with diesel over a wide range of con- the higher the molecular weight of the alcohol the less corrosive it
ditions, especially at low temperatures. Butanol, on the other hand, is [23]. The emission of toxic organic compounds such as aldehydes
has a higher heating value, higher cetane number, and lower vapor increases with the increase of alcohol content in the fuel mixture.
pressure. It is also less hydrophilic, and has a higher miscibility fac- However the minimum emissions of aldehydes correspond to the
tor in diesel than ethanol. The energy content of n-butanol fuel is stoichiometric composition of the fuel mixture and increase when
28.4 MJ/L and ethanol energy content is 21.2 MJ/L only. Because it is leaned or enriched with fuel. These can be reduced by adding
n-butanol is more hydrophobic than ethanol, it allows the use of to the fuel mixture small proportions of aniline and preheating the
currently existing fuel distribution infrastructure [13]. Moreover air before entering the engine [24]. The flash point of n-butanol is
ethanol is unsuitable for combustion in compression ignition en- quite high (30 °C) compared to ethanol (see Table 1) making it a
gines, because of its insufficient auto ignition quality [14]. Due to safer fuel to use than ethanol (flash point 13–14 °C).
high carbon content in alcohol molecules and good intersolubility Analysis of results and effects of the fuel properties on combus-
with diesel without addition of any co-solvents or emulsifiers, n- tion and emissions in reciprocating engines has been given some
butanol makes it easier to blend with diesel. An emulsifying agent attention by researchers. It is important to know the effects of fuel
is used to satisfy ethanol–gasoline fuel mixture homogeneity to properties on combustion and emissions generated so that it can
prevent phase separation [15]. The properties of n-butanol are be possible to effectively control the pollutants produced with lit-
such that n-butanol has the potential to overcome the drawback tle adjustments made on the engine. For instance, butanol is
brought about by low-carbon alcohols [6]. Butanol can be blended blended with vegetable oils to improve solubility of the vegetable
with diesel without phase separation [16]. Considering the supe- oils and reduce viscosity to aid flow ability [4].
rior qualities of butanol over ethanol, butanol would be preferable There is evidence in the literature that alcohol admixes have
to ethanol for blending with conventional diesel and more so than been successfully applied to conventional fuels to reduce regulated
methanol [15]. emissions. It is admissible that the use of smaller shared volume of
The properties of n-butanol, diesel, and ethanol are compared in alcohol with diesel is almost standard practice. The bioalcohols
Table 1. From the table the lower heating value, (LHV), and cetane that have similar or closer physical and chemical properties as die-
number are higher for n-butanol than ethanol [17]. While it is ar- sel are preferable. However, larger shared volumes of alcohol usu-
gued by others that alcohol extends the ignition delay in diesel en- ally mismatch some important properties (density, viscosity, and
gine [17], some researchers however pointed out that the solubility) that conventional fuels are able to meet and so limit
emissions in the exhaust gases are reduced. Different types of en- the bioalcohol application. It is important in the selection of biofu-
gines have been used for this purpose including two-cylinder, nat- els to select them in such a way that they do not affect significantly
urally aspirated or turbo-charged, light or heavy duty engines the properties that are necessary for engine stable operation.
[15,18–20]. Diesel fuel and B8, B16 and B24 mixtures [19] are alternately
The different structure of butanol isomers have a direct impact fired in a low load naturally aspirated, Ricardo Cusson diesel en-
on the physical properties of the biofuel [21]. Butanol (C4H9OH) is a gine [14,15]. Nitrogen oxides emissions increase from 400 to
higher-chain alcohol containing four carbon structures. It exists as 1200 ppm with increasing load from 1.4 to 5.37 bars as a function
different isomers based on the location of the hydroxyl (OH) group of brake mean effective pressure (BMEP). Diesel fuel indicates
and carbon chain structure. These are 1-butanol, also known as higher NOx concentration than the blends used. Properties of buta-
n-butanol, 2-butanol also known as sec-butanol, iso-butanol and nol fuel affecting these results include lower energy content of the
tert-butanol. The 1-butanol is a straight–chain structure with the biofuel given by LHV, higher heat of vaporization and lower cetane
alcohol (OH) at the terminal carbon. The 2-butanol has the hydro- number. Hydrocarbons that are generated by the blends (B8 and
xyl group at an internal carbon. Iso-butanol is a branched isomer B16) are higher than diesel.
with the OH group at the terminal carbon and tert butanol refers Carbon monoxide is reported to be lower [25] than at any other
to the branched isomer with the OH group at an internal carbon. load for the test fuels of neat diesel and blends (B8 than B16) at
Conventional fuel injection systems [22] usually encounter brake mean effective pressure (BMEP), 7.04 bars, for a bus diesel
lubrication problems due to the low viscosity of alcohol fuels, engine. Diesel fuel records the highest exhaust temperature at a
which leads to the wearing of the engine parts. During combustion, given operating load but only slightly higher than B8 and B16.
alcohol blended fuels produce acids that are responsible for wear- The blends indicate a higher hydrocarbon emission than DF when
ing of engine parts. Neutralizers such as zinc dialkyldithiophos- the load is increased. However, the hydrocarbon decreases when
 L. Siwale et al. / Fuel 107 (2013) 409–418 411

Table 1 2. Materials and methods

Properties of diesel, n-butanol and ethanol [15].

Fuel properties Diesel n-Butanol Ethanol 2.1. Experimental set-up and test procedure
fuel C4H9OH C2H5OH
Density at 20 °C (kg/m3) 837 810 788 Fig. 1 shows the schematic layout of the engine test bed and
Cetane number 50 25 8 instrumentation. The study was conducted on a four cylinder pis-
Lower heating value (MJ/kg) 43 33.1 26.8
ton, 1Z type, 1.9 L-66 kW Turbo-Direct Injection (TDI) Volkswagen
Kinematic viscosity at 20 °C 3.4a 3.6b 1.52b
(mm2/s) diesel engine. Typical fuel consumption data and brake specific fuel
Boiling point (°C) 180-360 118 78 consumption (BSFC) are shown in Table 2. The fuel injection tim-
Latent heat of evaporation (kJ/ 250 585 840 ings for the blends and DF controlled by the electronic diesel con-
kg) trol (EDC) for different speeds of 1500, 2500, 3000 and 3500 rpm
Oxygen (wt.%) 0 21.6 34.8
were 11°, 11°, 12° and 15° crank angle (CA) BTDC respectively at
Stoichiometric air–fuel ratio 15.0 11.2 9.0
Molecular weight 170 74 46 full load. The injection timing changed on part load in the function
a
of the load and the speed from 0° to 15° crank angle BTDC, but it
Ref. [31].
b was the same for the blends and DF.
Ref. [32].
The excess air ratio is given as k (see Table 2 for k values). The
engine performance and parameters correspond to the maps of
the shared volume of the alcohol with DF decreases. Exhaust gases
EDC. However, due to the lower energy content from the alterna-
in diesel engines during cold starting make a larger contribution to
tive fuels (blends), the brake power falls at full load, on part load
the total emissions than at any other operating condition of the en-
test, the BMEP were set to the same value. This causes changes
gine [26]. Smoky opacity for DF is higher than for blends. Smoky
in the value of k for the different fuel blends used (however the
opacity is higher at 1200 rpm than at 1500 rpm and is proportional
EDC calculation is more complex than here highlighted). Engine
to the load. Similar results in Refs. [17,27] are also obtained.
parameters are given in Table 3. The engine was fully equipped
The objective of this study was to evaluate the combustion
with a turbocharger with all the necessary sensors and actuators
characteristics and regulated emissions of neat DF and n-butanol/
for the stable operation of the engine. The engine was heated up
diesel blends (B5, B10, and B20 where B05 represent 5% shared vol-
until its cooling water temperature had reached 94 °C and was
ume of n-butanol with 95% diesel fuel) fired in a high load turbo-
kept at this constant temperature. Three blends were used: B5,
charged diesel engine. To build understanding of the impact of fuel
B10, and B20. Diesel D2 or B0 was used as the reference fuel (see
property on combustion and regulated emission, a similar study
Table 4). The engine ran for 20–30 min to warm up with either ref-
was used as a basis for comparison and attenuation of the results
erence fuel DF or test (blend) fuel. Once the new fuel was pumped
from this study.
in, the engine was made to run for another 20 min to allow for sta-
Fuel properties such as viscosity and solubility of additives and
ble operation of the new blend at test conditions. The study was
their mixture with diesel have a significant impact on the combus-
carried out for full, 75%, 50% and 25% load for the operating range
tion process and emissions produced. Although determination and
of engine speeds of 1500, 2500, 3000 and 3500 rpm. The engine for
evaluation of these performance variables was not intended, their
the test was also used by Lujaji et al. [28]. The engine was made to
underlying effects are used to interpret the results obtained in the
run for each measuring point on steady state condition for about
study.

Fig. 1. Schematic layout of engine test bed and instrumentation.

412 L. Siwale et al. / Fuel 107 (2013) 409–418

Table 2 tion according to ISO 9001. Torque was measured by a Borghi

Typical fuel consumption, k, BSFC and injection advance at 1500 rpm A = 100% and and Saveri FE-350S eddy current dynamometer, and the speed
B = 75% load and at 3000 rpm for DF and blends.
sensors placed on the dynamometer shaft. Some of the ancillary
DF B05 B10 B20 test equipment used is shown in Table 6. NOx concentration was
Fuel cons. (kg/h) measured by a Horiba 8120F analyzer using chemiluminescence
(A) at 1500 rpm 5.47 5.47 5.43 5.63 detector CLD-53M. The detector has measurement NO/NOx con-
(B) at 1500 rpm 3.0 3.06 3.12 3.12 centrations range between 100 ppm and 5000 ppm. The analyzer
(B) at 3000 rpm 7.13 7.34 7.56 7.62
module was equipped with a Horiba AIA-23 infrared gas analyzer
k () to measure carbon monoxide concentration. The unburned hydro-
(A) at 1500 rpm 1.4 1.42 1.5 1.52
(B) at 1500 rpm 1.72 1.82 1.92 1.92
carbon was measured by the hydrocarbon analyzer Horiba FIA-22
(B) at 3000 rpm 2.37 2.73 2.42 2.79 equipped with heated flame ionization detector. The smoke meter
BSFC (g/kW h)
AVL 415 (reflective type) was used to measure soot concentration
(A) at 1500 rpm 235 231 237 248 using a positive exhaust pressure line tapped from the engine ex-
(B) at 1500 rpm 252.9 258.02 263.09 263.08 haust outlet. The particulate matter was collected on a filter and
(B) at 3000 rpm 249.4 256.75 264.4 266.54 the meter measured smoke concentration on the base of changes
in reflectivity of the particulate matter and measured as Filter
Smoke Number (FSN).

Table 3
3. Theoretical estimations
Engine parameters.

Item Value/description The theoretical and empirical derivations, on which the experi-
Bore 79.5 mm mental emission results are based or can further be processed, are
Stroke 95.5 mm given in Eq. (1). The empirical formula for estimating soot concen-
Compression ratio 19.5
tration (PM) [6] is given as follows:
Maximum torque 202 Nm/1800 rpm
Maximum power 66 kW/4000 rpm
soot ¼ 1=0:405  5:32  FSN  e0:3062FSN  0:001  ðmair
Fuel system
Injector pump Distributor-type (Bosh VP37) þ mfuel Þ  1=1:2929 ð1Þ
Combustion chamber Bowl in piston
Injector type and pressure 5 Hole, 180 bars The Filter Smoke Number FSN (dimensionless) is registered by
Injection timing 0–15 deg BTDC an AVL 415 meter. In Eq. (1) mair denotes air mass flow rate (kg/
h) and mfuel denotes fuel mass flow rate (kg/h). From this formula
(when divided by engine operating power) the specific dry soot
emission (g/kW h) can be estimated. This empirical formula is valid
Table 4
for the range of (0–8.08) FSN. The lower heating value of blend is
n-Butanol/diesel mixture ratio by volume.
calculated from following equation:
Item Alcohol % Base fuel % ,
X
2
vi X2
vi
Blend B5 n-Butanol 5 Diesel (D2) 95 LHV bl ¼  qi  LHV i  qi ð2Þ
Blend B10 n-Butanol 10 Diesel 90
1
v i¼1
v
Blend B20 n-Butanol 20 Diesel 80
where LHVbl is the lower heating value (MJ/kg) of the blend; LHVi is
the lower heating value of component i of the blend, i = 1 is n-buta-
nol, i = 2 is diesel. Eq. (3) is used to calculate density of the blends
Table 5 [29]:
Operating conditions of the engine for DF.
a X
2
vi
Torque (Nm) Speed (rpm) Exhaust Temperature (°C) MAP, bar q¼  qi ð3Þ
184.6 3004 562 0.82 i¼1
v
136.4 3004 473 0.71
92 3002 419 0.56 where q is the density of the blend (kg/m3) and vi/v is the volume
45 3000 333 0.38 fraction of component i of the blend. Eq. (4) is applied for the calcu-
168.9 3504 584 0.79 lation of air-to-fuel ratio,
187.8 2505 519 0.84 ( ),
153.7 1502 532 0.35 X2
vi X2
vi
A=Fsbl ¼  qi  A=Fsi  qi ð4Þ
a
Manifold (boost) air pressure (rel.); ambient temp = 20 °C, intake air i¼1
v i¼1
v
temperature = 25 °C.
where A/Fsbl is the stoichiometric air fuel ratio of the blend and A/Fsi
is the stoichiometric air fuel ratio for i component of the blend. The
2 min before recording the values. The engine operating conditions
calculated properties of all the test blends are shown in Table 7. The
are shown in Table 5. The in-cylinder pressure was measured by a
heat release rate is determined by the adiabatic energy balance (5)
piezoelectric pressure transducer Kistler KIAG 600. The Top Dead
reported by Lujaji et al. [28]:
Center (TDC) crank angle position was measured by an Optical En-
coder HENGSTLER RI 32-0/1024.ER.14KA.  
dQ 1 dv dp
Fuel consumption was measured using the AVL 7030 dynamic ¼ c p þV ð5Þ
dh c  1 dh dh
fuel consumption measuring equipment. The fuel balance works
on the gravimetric measuring principle. Fuel is supplied to the en- The term on the left hand side dQdh
is the heat release rate. In Eq.
gine from a measuring vessel inside the instrument where the (5) c is the specific heat ratio, and h is crank angle. On the right
weight of the fuel is continuously measured. This instrument hand, V is the instantaneous engine displacement, and dp dh
is the
enables the highest temperature stability of the fuel conditioning change in indicated pressure in the combustion chamber divided
system with measuring accuracy of 0.12%; including self-calibra- by the crank angle change and ddhv is the change in volume divided
 L. Siwale et al. / Fuel 107 (2013) 409–418 413

Table 6 volume 30% (v/v) or BD30 compounded with diesel fuel (DF) with
Ancillary equipment. properties similar to DF used in our study. A third admix to BD30
Item Type tested is ethanol 5% and 7.5% (E5 and E7.5). The brake specific fuel
Torque measurement by eddy FE350S-Borghi and Saveri (moment of consumption (BSFC) for the admixed blends B30 + 5E and
current dynamometer inertia, 0.618 kg/m2) B30 + 7.5E is similar with BSFC values for DF. For the experiments
Pressure transducer Kistler KIAG 600 in study A the power rating for the engine is 55 kW, compression
TDC and crank angle speed pick up Optical encoder HENGSTLER RI 32-0/ ratio (CR) is 16.5:1 and speed was kept constant at 1600 rpm.
1024.ER.14KA
Thermocouple K type
The blends BD30 + 5E and BD30 + 7.5E indicate lower NOx emis-
Fuel measurement AVL 7030 sions than pure BD30. The significance of study A is that small
quantities of bioethanol eliminate increase of NOx caused by BD
(relative to DF).
Nitrogen oxides concentration increased as the load increased,
Table 7 showing more or less similar trends for test fuels at 1500, and
Calculated properties of test blends.
3000 rpm and at 25%, 50% and 75% and full load. The blends
Property Diesel n-Butanol B5 B10 B20 showed increased NOx concentration with increased shared vol-
Density (kg/m3) 0.86 0.81 0.858 0.855 0.85 ume of n-butanol than DF at 75% load, 1500 rpm. For 100% load
LHV (MJ/kg) 42.5 35.1 42.15 41.8 41.08 at 1500 rpm, NOx concentration for B10, B20 and DF were similar
Stoichiometric A/F ratio 15 11.21 14.15 14.01 13.71 while B5 reduced slightly; the differences are within the range of
Latent heat (kJ/kg) 250 585 284.71 299.5 329.3
accuracy of the analyzer. The reason for these differences can be
explained as follows. In the function of blending ratio or shared
volume of n-butanol with DF, the cetane number and lower heat-
Table 8 ing (calorific) value decrease as DF is substituted. As can be seen
Diesel fuel and n-butanol specifications. in Table 1, the latent heat of evaporation of n-butanol is higher
Item Diesel fuel (D2) n-Butanol than DF, so the blends caused a decrease in NOx emission (due to
Molecular formula H3(CH2)2CH2OH evaporative cooling), but the oxygenated fuel components in-
Molecular weight (kg/kmol) 74.12 creased the combustion efficiency, and the premixed combustion
Purity (%) 99.99 became more intensive, causing an increase of NOx emission. These
Specific density (g/cm2) P0.82 at 15 (°C) 0.809 two tendencies kept the NOx emission on the same level in the
Boiling point (°C) at 1013 hPa 180–360 118
function of blending ratio, except NOx emission increased by
Melting point (°C) 89.8
Flash point (°C) >55 (°C) 30 10.3%, 32.3%, and 54.4% above DF at 75% load and speed
Cetane number P51 1500 rpm for B5, 10 and 20 respectively. This is probably due to
Water content (mg/kg) 6200 the effect of the working process of the EDC.
Kinematic viscosity (mm2/s) 2.0–4.5 at 40 (°C)
Fig. 3a shows the effect of blends on exhaust gases temperature
Sulfur content 610 (mg/kg)
(EGT) and Fig. 3b shows the effect of blends on torque at different
speeds at full load. Due to the lower energy content of the blends
the released heat is lower for the same injected dose which leads
by the crank angle change. Using a computer model the heat to reduced torque and EGT. The higher latent heat of evaporation
release was evaluated using the pressure values from the computer of the fuel also reduces that temperature, also reported by Dogan
pressure traces and the instantaneous volume evaluated from the [16]. The highest EGT and torque for every measured speed was
following equation: shown by DF followed by B05, B10 and B20.
8 9 Fig. 4 shows the effect of blends on CO concentration at speeds
>   >
dv pA < ph R2 sin ð180
2 ph
Þ = 1500 and 3000 rpm. Carbon monoxide is formed due to the incom-
¼  r sin þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð6Þ
dh 180 > : 180 > plete combustion of all the hydrocarbon compounds. In study A, CO
2  ð1  R2 sin ð ph ÞÞ;
2
180 measured is as follows. For DF: At BMEP = 0.650, 0.325, 0.1 (MPa)
CO = 912, 351, 634 ppm respectively. In our study the results were
where A is the cylinder cross sectional area, R = l/r connecting rod
as follows. DF: At BMEP = 0.751, 0.50, 0.251 (MPa) CO = 286, 355,
length to radius ratio, h crank angle.
334 ppm, respectively. In study A for B30 + E5  7.5% at BMEP =
0.650, 0.325, 0.100 (MPa) CO = 300, 575, 280 ppm respectively. In
4. Materials our study for B20, at BMEP = 0.751, 0.50, 0.251 (MPa) CO = 111,
234, 338 ppm respectively. It is observed that using n-butanol
The fuel n-butanol was manufactured by VWR Prolabo (BDH) blended with DF in a turbo charged engine CO is substantially re-
and Diesel used for the experiments was D2 standard EN 590 as duced as opposed to the use of BD30 plus 5–7.5% bioethanol in a
shown in Table 8. naturally aspirated engine. In both cases using biofuels reduced
CO generally relative to DF.
5. Results and discussion Carbon monoxide concentration emission increased at lower
speed of 1500 rpm than at 3000 rpm for the tested fuels. In the
Fig. 2 shows the effect of blends on emission of nitrogen oxides presence of an abundant oxidant (air), more complete oxidation
(NOx) at speeds 1500 and 3000 rpm at different loads. Thermal NOx of the fuel results towards the formation of carbon dioxide rather
is a by-product of combustion and its formation rate has exponen- than CO [12]. Besides an air-deficient local environment (fuel-rich)
tial dependence on temperature. This means that its probable for- generates high CO concentration. At low load the fuel lean environ-
mation is more intensive with higher temperatures. Thermal NOx ment absorbs energy from the combustion products thereby
formation which is well studied and understood [12] depends on reducing the temperature causing the CO oxidation to slowdown.
the residence time that combustion gases stay on high temperature. The increased injected fuel at high loads decreases the overall ex-
Raslavičius and Bazaras [12] studied (we call it study A) multi cess air ratio and increases the CO concentration.
component fueling of a naturally aspirated diesel engine using The blends indicated lower emission of CO than DF at 1500 rpm
biodiesel (BD) from rapeseed oil methyl esters (RMEs) of shared above 50% load. Due to higher oxygen content in the blends more
414 L. Siwale et al. / Fuel 107 (2013) 409–418

Fig. 2. NOx emissions for different loads at (a) 1500 rpm at (b) 3000 rpm.

Fig. 3. Effect of blends on (a) exhaust gases temperature and (b) torque at different speeds at full load.

of the CO is oxidized to CO2 than when DF only is burned. All fuels (MPa) UHC = 14, 20, 30 respectively and for B20, UHC = 30, 37,
had almost similar CO concentration at 75% and 50% at 3000 rpm 55 ppm for the same range of BMEP. As was expected the UHC in-
but lower at 75% than at 50% load. Increasing the fraction of the creased over DF due to the longer ignition delay of the n-butanol/
alcohol content in the blends, B05, B10 and B20 reduce the CO DF blends in our study.
emissions by 35.7%, 57.1% and 71.4% respectively at 75% load, Unburned hydrocarbon emission concentration was generally
1500 rpm. Due to the lower stoichiometric air–fuel ratio of n-buta- higher for n-butanol/DF blends than DF at different torques. How-
nol and the same injected dose of the fuels the excess air ratio is ever at 3000 rpm the increase of the emission is much smaller over
significantly higher at 100% load for the blends decreasing the CO 25% load. Reasons given for the observed increase of UHC emission
emission intensively. At 25% load the CO emission increases for include slower evaporation, and higher heat of evaporation of n-
the blends especially at 3000 rpm. The reason for this is mainly butanol blends. The increase could also be due to a phenomenon
that the higher latent heat of the blends generates cooler environ- which is known as ‘‘lean-outer-flame-zone’’. Beyond this zone the
ment for combustion. fuel is already mixed and could not be ignited, because it is out of
Fig. 5 shows the effect of the blends on the unburned hydrocar- the lean flammability limit [30]. This phenomenon is more signifi-
bon emission concentration at 1500 and 3000 rpm. The amount of cant as the cetane number decreases. If the fuel–air mixing is im-
fuel injected during the ignition delay period has an influence on proved and the in-cylinder temperature increases as is the case at
the concentration of the unburned hydrocarbon. For BD30 in study 3000 rpm at higher loads, the unburned hydrocarbon is generally
A with higher cetane number (CN) [12] the injection period is short- reduced; this effect reduces the influences of the n-butanol blends.
er decreasing correspondingly the emission of the UHC. The CN for Unburned hydrocarbons increased relative to DF at 75% load at
n-butanol/DF blend is lower than RME and therefore the injection 1500 rpm by 21.4%, 71.4%, and 214% for B5, B10, and B20 respec-
period for n-butanol blends would be longer during the ignition de- tively but reduced at 75%, 100% loads at 3000 rpm to below 20 ppm.
lay period expecting a higher UHC emission. In study A for DF: at Fig. 6 shows the soot content in the exhaust gas at 1500 and
BMEP = 0.650, 0.325, 0.100 (MPa) UHC = 318, 347, 406 ppm respec- 3000 rpm. In a direct injection diesel engine, the fuel–air distribu-
tively. For B30 + 5  7.5E UHC = 220, 240, 280 ppm for the same tion is not homogeneous. This means that soot formation is con-
range of BMEP. In our study for DF: BMEP = 0.751, 0.50, 0.251 trolled by mixing of air and fuel besides excess air ratio [12].
 L. Siwale et al. / Fuel 107 (2013) 409–418 415

Fig. 4. Carbon monoxide emission for different loads at (a) 1500 rpm at (b) 3000 rpm.

Fig. 5. UHC emission for different loads (a) at 1500 rpm (b) at 3000 rpm.

Soot formation depends on the condition of the engine, type of the bustion process and eventually the levels of regulated emissions
combustion chamber and differences in the physic chemical produced from the engines.
properties of fuel [12]. Blending n-butanol with DF reduces soot, In study A using BD30 alone shortens the ignition delay (ID).
because the bonded oxygen atoms from the hydroxyl group of Adding 10% ethanol (BD30 + E10) is found to prolong ID which is
n-butanol reduces soot formation mainly by moderating precursor not desirable as it results in hard start of the test engine. Compar-
production. The low C/H ratio of n-butanol also reduces soot for- ing cetane numbers, rapeseed oil esters (RME) is 55 and DF used is
mation. In study A, as well as in this study, soot concentration 51. Cetane number for n-butanol is approximately 25 and ethanol
showed substantial reduction. Soot concentration reduced when CN is approximately 8.
n-butanol/DF blends were used as explained in the preceding text. Blending greater than 20% (v/v) shared volume of n-butanol
At both speeds 1500 and 3000 rpm, soot concentration for the with DF reduces further the cetane number of the mixture. There-
blends is always less at a given load than for DF. Relative to DF, soot fore the limiting point would be when the CN of the mixture is too
emission reduced with increasing shared volume of n-butanol in low for the engine to start. By this short analysis it is therefore not
the blend by 55.5%, 77.8%, and 85.1% for B5, 10, 20 respectively advisable to exceed the ignitability level of the fuel mixture. So a
at 75% load and 1500 rpm. At 3000 rpm soot emission was reduced shared volume of n-butanol with DF, higher than the limit cannot
further for all the test fuels from the values at 1500 rpm. At 75% be permitted because a reduction in CN of the mixture leads to a
load soot concentration reduction was highest. The little deviance hard start of engine. Since CN for RME is higher than n-butanol,
from the trend at 25% and 50% load at 3000 rpm is not well known: 20% or slightly more than this shared volume of n-butanol may
it might be caused by the characteristics of temperature of the not be exceeded. Adding 5–20% n-butanol fraction to diesel in
combustion process, as temperature also plays an important role the blend preparation stage not only increases the ignition delay
in soot formation and oxidation. progressively during engine operation according to the fraction
In combustion studies the effects of fuel property denoted by added but also reduces the combustion duration. Combustion
cetane number measures the easiness that a fuel ignites. The igni- duration is reduced due to the improved combustion efficiency ex-
tion delay is a time lag that is stipulated by the physical chemical pected by high n-butanol fractions. The ignition delay is slightly
properties of the fuel. It is measured in time (s) or crank angle de- prolonged due to the displaced DF which subtracted higher CN
grees in engines cycles. This important fuel quality affects the com- components from the blend than n-butanol did add. These results
416 L. Siwale et al. / Fuel 107 (2013) 409–418

Fig. 6. Soot emission for different loads (a) at 1500 rpm (b) at 3000 rpm.

Fig. 7. Indicated pressure and the heat release rate vs crank angle at 2500 rpm.

Fig. 8. Indicated pressure and heat release rate vs crank angle at 3000 rpm.

agree with findings in study A [12] who obtain a prolonged ignition to-air mixture prior to ignition caused a rapid occurrence of the
delay by adding the alcoholic admix of E10 (10% ethanol by vol- premixed combustion phase (see Figs. 7–9).
ume) to BD30 and is shortened by reducing the alcohol fraction. Figs. 7–9 show the indicated pressure and heat release rate at
In this study it was observed that the amount of prepared fuel- 2500, 3000, and 3500 rpm for the entire test fuels at full load.
 L. Siwale et al. / Fuel 107 (2013) 409–418 417

Fig. 9. Indicated pressure and the heat release rate vs crank angle at 3500 rpm.

stable combustion characteristic than the reference fuel. B10

showed less stable combustion quality by indicating a higher stan-
dard deviation at all speeds than the other blends.

6. Conclusions

A comparison of regulated emissions in this study using a tur-

bo-charged diesel engine with a similar study in literature [12]
(based on a naturally aspirated diesel engine) has shown an
improvement with a turbo-charged engine in the reduction of
NOx, CO, UHC and soot.
The reduction in NOx and CO using n-butanol/DF blends: B05,
B10, and B20 (where B05 stands for 5% shared volume of n-buta-
nol) in this study was more substantial than in a similar study
using 30% shared volume of rapeseed oil ethyl esters (RMEs) with
DF admixed to 5% and 7.5% bioethanol [12]. In both cases no mod-
ification to the engine was required. NOx emission increased in this
Fig. 10. Standard deviation of pressure cycles at different speeds.
study by 10.3%, 32.3%, and 54.4% above DF at 75% load at speed
1500 rpm for B5, 10 and 20 respectively. Increasing the fraction
Although there was little difference in the pressure curves due to of the alcohol content in the blends, B05, B10 and B20 reduced
the blends as observed in Figs. 7–9, increased premixed heat re- the CO emissions by 35.7%, 57.1% and 71.4% respectively at 75%
leases was observed due to the blends. Notwithstanding an in- load at 1500 rpm.
crease of heat release of the test fuels at lower speed, there was Unburned hydrocarbons are reduced significantly using n-buta-
no differentiation of the peak pressure. With a reduction of speed nol/DF blends in this study than 30% RME and 5% to 7.5% bioetha-
from 3500 to 2500 rpm the peak-heat release rate increased as nol in the other study [12]. However, the blends produce higher
the residence time in the combustion chamber increased with UHC than DF in both cases. In this study the UHC emission concen-
reducing speed. Figs. 7–9 present the increased heat release rate tration increases by 21.4%, 71.4%, and 214% over DF at 75% load at
in the crank angle region, h = 350–360°, indicating the event of pre- 1500 rpm for B5, B10, and B20 respectively. From the foregoing
mixed combustion. As a possible explanation of the formation of discussion n-butanol in smaller shared volume up to 20% (v/v)
the premixed fuel portion, the following mechanism can be consid- with DF has shown to be a highly recommended additive to diesel
ered. The cetane number of the blends decrease with the addition fuel especially on a turbo-charged engine needing no modification
of n-butanol and the boling point of n-butanol is lower. The mixing to the engine if a substantial reduction in the regulated emissions
ability of the blend was enhanced as n-butanol has higher oxygen is to be made.
content. Therefore, the heat release rate in the premixed phase in- Soot emission was greatly reduced in both compared studies (in
creased with addition of n-butanol ratio. In the mixing controlled this study and Ref. [12]). This was when bioethanol admix was
combustion phase, it was observed that the heat release rate for introduced to BD30/DF blend in the other study [12]. In both cases
the blends showed a slightly steeper profile of the heat curve just smaller proportions of bioalcohol was used. In this study the soot
after the premixed combustion phase than DF. By this the combus- reduction relative to DF at 1500 rpm at 75% load for B05, B10
tion duration of the blends was slightly shortened in comparison to and B20 mixtures was 55.5, 77.8 and 85.1% respectively. This
DF as a result of a better combustion efficiency of the blends. reduction is a significant advantage of blending DF with smaller
Fig. 10 shows the standard deviation of the pressure cycles. It shared volumes of bioalcohol.
can be deduced that blends have a less pressure deviation from A distinguishable and amplified premixed combustion phase
the average value than diesel. In other words blends have a more was observed with increased n-butanol/DF shared volume with
418 L. Siwale et al. / Fuel 107 (2013) 409–418

DF. This suggested an increased effectiveness of the air–fuel mixing [10] Dernotte J et al. Evaluation of butanol-gasoline blends in a port fuel-injection,
spark-ignition engine. Oil Gas Sci. Technol. Rev. IFP 2010;65(2):345–51.
during the ignition delay period. There were no significant differ-
[11] Valentino G et al. Experimental study on performance and emissions of a high
ences in the indicated pressure between DF and the n-butanol/DF speed diesel engine fuelled with n-butanol diesel blends under premixed low
blends although the heat release increased with reduced speed. A temperature combustion. Fuel 2012;92(1):295–307.
higher value of the standard deviation for DF than the blends [12] Raslavičius L, Bazaras Z. Variations in oxygenated blend composition to meet
energy and combustion characteristics very similar to the diesel fuel. Fuel
was observed from the standard deviation diagram, indicating a Process Technol 2010;91(9):1049–54.
more stable combustion process for the blends than DF. [13] Bryan W et al. Autoignition of n-butanol at elevated pressure and low-to-
intermediate temperature. Combust Flame 2011;158:809–19.
[14] Murat C, Yuksel F, Hus H. Emission characteristics of a converted diesel engine
Acknowledgments using ethanol as fuel. Energy Sustain Dev 2009;13:250–4.
[15] Rakopoulos DC et al. Effects of butanol-diesel fuel blends on the performance
The authors acknowledge and are greatly indebted for the and emissions of a high-speed DI diesel engine. Energy Convers Manage
2010;51(10):1989–97.
financial support from the joint research collaboration between [16] Dogan O. The influence of n-butanol/diesel fuel blends utilization on a small
Hungary/South Africa Funding (UID 72384 and TÉT_10-1-2011- diesel engine performance and emissions. Fuel 2011;90(7):2467–72.
0005); for facilitation by the two universities, Tshwane University [17] Karabektas M, Hosoz M. Performance and emission characteristics of a diesel
engine using isobutanol-diesel fuel blends. Renewable Energy
of Technology, Pretoria, South Africa and Budapest University of 2009;34(6):1554–9.
Technology and Economics, Budapest, Hungary; the latter for mak- [18] Broustail G et al. Experimental determination of laminar burning velocity for
ing available the laboratory facility, Jendrassik Gyorgy hotechnikai butanol and ethanol iso-octane blends. Fuel 2011;90(1):1–6.
[19] Veloo PS et al. A comparative experimental and computational study of
Laboratorium in Budapest BME, where the engine experiments
methanol, ethanol, and n-butanol flames. Combust Flame
were conducted. This work is connected to the scientific program 2010;157:1989–2004.
of the ‘‘Development of quality-oriented and harmonized R+D+I [20] Zoldy M, Hollo A, Thernesz A, Butanol as diesel extender option for internal
strategy and functional model at BME’’ Project. This Project is sup- combustion engine. SAE World Congress, Paper Number: 10SDP-0015; 2010.
[21] Wallner T, Miers SA, McConnell S. A comparison of ethanol and butanol as
ported by the New Szechenyi Development Plan (Project ID: oxygenates using a direct injection, spark ignition engines. Eng Gas Turbines
TÁMOP-4.2.1/B-09/1/KMR-2010-0002). The authors are grateful Pow 2009;131:0328021–29.
to the support of Bólyai János Research fellowship of HAS (Hungar- [22] Surisetty V, Rao D, Ajay K, Kozinski J. Alcohols as alternative fuels: an
overview. Appl Catal A 2011;404(1–2):1–11.
ian Academy of Science). Authors are grateful for the support of [23] Rasskazchikova TV, Kapustin VM, Karpov SA. Ethanol as high–octane additive
Prof. Dr. Florian Heinitz, Director of Transport, and Spatial Planning to automotive gasolines. production and use in Russia and abroad. Chem
Institute in University of Applied Science in Erfurt, Germany. Technol Fuels Oils 2004;40(4):203–10.
[24] Karpov S. Aliphatic alcohols as components of fuels for internal combustion
engines. Chem Technol Fuels Oils 2008;44(4):249–54.
References [25] Rakopoulos DC et al. Investigation of the performance and emissions of bus
engine operating on butanol/diesel fuel blends. Fuel 2010;89(10):2781–90.
[1] Zhang Y, Boehman AL. Oxidation of 1-butanol and a mixture of n-heptane/1- [26] Rakopoulos CD et al. Study of turbocharged diesel engine operation, pollutant
butanol in a motored engine. Combust Flame 2010;157:1816–24. emissions and combustion noise radiation during starting with bio-diesel or n-
[2] Sarathy SM et al. An experimental and kinetic modeling study of n-butanol butanol diesel fuel blends. Appl Energy 2011;88:3905–26.
combustion. Combust Flame 2009;156(4):852–64. [27] Xiaolu Li XQ, Zhang L, Fang J. Combustion and emission characteristics of a
[3] Demirbas FM, Balat M, Balat H. Biowastes-to-biofuels. Energy Convers Manage two-stroke diesel engine operating on alcohol. Renewable Energy
2011;52(4):1815–28. 2005;30:2075–84.
[4] Rakopoulos DC et al. Combustion heat release analysis of ethanol or n-butanol [28] Lujaji F et al. Experimental investigation of fuel properties, engine
diesel fuel blends in heavy-duty DI diesel engine. Fuel 2011;90(5):1855–67. performance, combustion and emissions of blends containing croton oil,
[5] Thomas SCE. Transportation options in a carbon-constrained world: hybrids, butanol, and diesel on a CI engine. Fuel 2011;90(2):505–10.
plug-in hybrids, biofuels, fuel cell electric vehicles, and battery electric [29] Yusuf T, Buttsworth D, Nagafi G. Theoretical and experimental investigation of
vehicles. Int J Hydrogen Energy 2009;34(23):9279–96. SI engine performance and exhaust emissions using ethanol–gasoline blended
[6] Yao M et al. Experimental study of n-butanol additive and multi-injection on fuels. ICEE 2009;3:195–201.
HD diesel engine performance and emissions. Fuel 2010;89(9):2191–201. [30] Rakopoulos DC et al. Investigation of the performance and emissions of bus
[7] Szwaja S, Naber JD. Combustion of n-butanol in a spark-ignition IC engine. Fuel engine operating on butanol/diesel fuel blends. Fuel 2010;89(10):2781–90.
2010;89(7):1573–82. [31] Liu Y, Jiao W, Qi G. Preparation and properties of methanol–diesel oil
[8] D’urre P. Fermentative butanol production: bulk chemical and biofuel. Ann. emulsified fuel under high-gravity environment. Renewable Energy
NY. Acad. Sci. 2008;1125:353–62. 2011;36(5):1463–8.
[9] Singliar M, Dolnak M, Emerging technologies for second generation of biofuels. [32] Pukalskas S et al. The Mixture of biobutanol and petrol for Otto engines.
MOL scientific Magazine, SD and HSE special; 2008. p. 67–80. Transport 2009;24(4):301–7.