You are on page 1of 10

Energy 88 (2015) 101e110

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

A comprehensive analysis of the effect of ethanol, methane and


methane-hydrogen blend on the combustion process in a PFI (port
fuel injection) engine
F. Catapano a, S. Di Iorio a, *, A. Magno a, b, P. Sementa a, B.M. Vaglieco a
a
Istituto Motori e CNR, Napoli, Italy
b
Universita Federico II, Napoli, Italy

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

Article history: The energy crisis and environmental issues make the alternative fuels, both liquid and gaseous, even
Received 3 November 2014 more attractive because of their potentiality in reducing the fuel consumption and the pollutant emis-
Received in revised form sions. Ethanol is the most promising alternative liquid fuel for spark ignition engines. It has a higher
2 February 2015
octane number, which provides good anti-knock characteristics and in the possibility to work with
Accepted 16 February 2015
Available online 20 March 2015
higher compression ratios, so improving the engine efciency. The higher heat of vaporization compared
to gasoline leads to an increased power output. Moreover, the larger oxygen content provides a more
complete combustion and therefore reduced emissions. Among gaseous fuels, methane is considered one
Keywords:
Ethanol
of the most interesting. It has wider ammable limits and better anti-knock properties than gasoline.
Methane Moreover, it is characterized by lower CO2 emissions. On the other hand, the slow ame propagation
Hydrogen speed and its poor lean-burn capability produce lower engine power output with respect to gasoline. The
Energy sustainability addition of a high burning velocity fuel, such as hydrogen, allows to improve the combustion process in
Internal combustion engines terms of burning velocity and extend the lean operation limit. The objective of this paper is the analysis
of the effect of different fuels on the engine performance and emissions. Experimental investigations
were carried out in an optically accessible small single-cylinder, spark ignition four-stroke engine. It was
equipped with the cylinder head of a Port Fuel Injection commercial 244 cc engine. The engine was
fueled with gasoline, ethanol, methane and a blend of hydrogen in methane. Optical measurements were
performed to analyze the combustion process with high spatial and temporal resolution. In particular,
the optical techniques based on 2D-digital imaging were used to follow the ame propagation in the
combustion chamber. UVevisible spectroscopy allows the detection of chemical markers of the com-
bustion process such as the radicals OH* and CH*. The exhaust emissions were characterized by means of
gaseous analyzers. The measurements were performed at steady state conditions.
2015 Elsevier Ltd. All rights reserved.

1. Introduction The alcohol biofuels are considered competitive alternative fuels


because of their potentiality to replace petroleum fuels and reduce
The growing fossil fuels reserves depletion and the increasing particle emissions [1e4]. They can be produced from biomass re-
concerns on environmental pollution problems associated to the sources such as natural products or waste materials. The biomass
internal combustion engines have led to a renewed interest on represents an endless source of feedstock for the production of
clean alternative fuels obtained from the renewable sources. These biofuels. Ethanol is the most promising alternative fuel for spark
fuels have the potentiality to reduce the exhaust emissions, ignition engines [5e8]. It has a higher octane number with respect
allowing the engine to comply with the more restrictive standard to gasoline, which provides good anti-knock characteristics. This
emissions regulations [1] and contribute to energy sustainability. property makes it suitable to work at higher compression ratios,
thus improving the engine efciency. It has a higher heat of
vaporization compared to gasoline, producing a stronger charge
cooling effect. Consequently, a large mass of air is drawn into the
* Corresponding author. Istituto Motori e CNR, Via G. Marconi 4, 80125 Napoli,
cylinder increasing the volumetric efciency. Moreover, the larger
Italy. Tel.: 39 0817177225; fax: 39 0812396097.
E-mail address: s.diiorio@im.cnr.it (S. Di Iorio). oxygen content provides a more complete and cleaner combustion

http://dx.doi.org/10.1016/j.energy.2015.02.051
0360-5442/ 2015 Elsevier Ltd. All rights reserved.
102 F. Catapano et al. / Energy 88 (2015) 101e110

that produces a lower rate of CO (carbon monoxide), HC (hydro-


carbon) and particulate emissions [5e8]. Among the gaseous fuels,
methane is considered the most favorable for the internal com-
bustion engines because of its physical and chemical features [9]. It
shows an immediate advantage over other hydrocarbon fuels
because of the lower C/H ratio, which produces reduced CO2 (car-
bon dioxide) emissions. The main drawbacks of methane are the
slow burning velocity, the poor lean-burn capability and the air
displacement, which causes cycle-by-cycle variations, lower engine
power output and large fuel consumption [10]. The slow burning
velocity is improved by mixing methane with a fast burning ve-
locity fuel, typically hydrogen. Hydrogen is an excellent additive as
its laminar burning velocity is seven times higher than methane
[11e13]. Moreover, hydrogen is characterized by a more stable
combustion due to a wider combustion limit and low ignition en-
ergy, thus allowing the extension of the lean operation limit
[14e18]. A comprehensive analysis of the effect of alternative fuels
on combustion process is essential to improve properly the use of
the fuel, contributing to the energy consumption reduction. The
paper aims to characterize the effects of different fuels on the Fig. 1. Optical engine.

combustion process and exhaust emissions. The measurements


were performed on a transparent small displacement single cyl-
and the injection duration, were set by means of a programmable
inder, spark ignition engine. It was fueled with gasoline, ethanol,
electronic unit.
methane and a blend of hydrogen at 20% v/v in methane (M20H2).
An optical shaft encoder was used to transmit the crank shaft
Tests were carried out at 2000 rpm at low and full load. A combi-
position to the electronic control unit. The information was in
nation of optical techniques was applied to analyze the combustion
digital pulses, the encoder has two outputs, the rst is TDC (top
process with high spatial and temporal resolution. In particular,
dead center) index signal, and it has a resolution of 1 pulse/revo-
high-speed 2D-digital imaging measurement in the UVevisible
lution. The second is the CDM (crank angle degree marker) 1 pulse/
range was performed. This technique is widely used to investigate
0.2 . The engine is a 4-stroke and the encoder gives as output two
the ame propagation. Moreover, a detailed analysis of the effect of
TDC signals per engine cycle so, in order to have the right crank
the fuels on the combustion process was carried out by means of
shaft position, one pulse is suppressed via software.
the ame emission spectroscopy in the UVevisible range. The
A quartz pressure transducer was ush installed in order to
chemiluminescence emission from OH* and CH* were measured to
measure the in-cylinder pressure with a sensitivity of 19 pc/bar and
assess the combustion process from quality perspective. The
a natural frequency of 130 kHz. The in-cylinder pressure, the rate of
gaseous and particulate matter emissions were measured at the
heat release and the related parameters were assessed on an in-
exhaust by means of conventional instruments.
dividual cycle basis and/or averaged on 400 cycles. The perfor-
mance in terms of the IMEP (indicated mean effective pressure) and
2. Experimental apparatus and procedure
the CoV (coefcient of variance) of IMEP was assessed as well.
The experimental apparatus included the following modules:
the optically accessible PFI (port fuel injection) spark ignition en- 2.2. Experimental setup for optical investigation
gine, an electrical dynamometer, the fuel injection line, the data
acquisition and control units as well as the emissions measurement The conguration of the experimental apparatus used for the
systems. optical investigations is shown in Fig. 2.
During the combustion process, the light passed through the
2.1. Engine sapphire window and was reected toward the optical detection
assembly by a 45 inclined UVevisible mirror located in bottom of
The engine is a transparent small displacement single cylinder the engine. Spectroscopic measurements were carried out by
(Fig. 1). The main engine specications are reported in Table 1. means of a spectrograph coupled to an ICCD (intensied charge
The head had four valves and a centrally located spark plug. A coupled device) camera. The chemiluminescence signals were
four-valve, pent-roof chamber engine was tted on an elongated detected in the central locations of the combustion chamber. The
piston. The engine was characterized by an elongated cylinder and light from the sapphire window was focused by a 78 mm focal
a piston provided with a sapphire window, which replaced the at- length, f/3.8 UV Nikon objective onto the inlet slit of the spectro-
bottom piston bowl. The engine was also equipped with a quartz graph. The slit was 250 mm wide open. The spectrograph had a
cylinder in order to have a lateral view of the combustion chamber.
Table 1
This system enabled the passage of optical signals coming from the Engine specications.
combustion chamber. To reduce the window contamination by
Engine PFI spark ignition
lubricating oil, the elongated piston arrangement was used
together with self-lubricating Teon-bronze composite piston rings Bore [mm] 72
Stroke [mm] 60
in the optical section. The engine reached a maximum speed of
Connecting road [mm] 130
5000 rpm. Displacement [cm3] 244
The air fuel ratio was measured by a linear lambda sensor Clearance volume [cm3] 25.7
installed at the exhaust. A closed loop control allowed the adjust- Compression ratio 10.5:1
ment of the fuel injection in order to work with the desired air fuel Maximum power [kW] 7.9 @ 5000 rpm
Maximum torque [N m] 14.7 @ 5000 rpm
ratio value. The injection parameters, such as the ignition timing
F. Catapano et al. / Energy 88 (2015) 101e110 103

Fig. 2. Optical setup for spectroscopic measurements and digital imaging.

150 mm focal length with a 300 line/mm diffraction grating. The statistics based on 250 cycles was obtained. In this way, a good
spectral image formed on the spectrograph exit plane was matched accuracy on the R trend was achieved. In particular, in the worst
with an ICCD camera. Data were detected with the spectrograph case in terms of cycle-to-cycle variability, the coefcient of deter-
placed at two central wavelengths, 325 and 575 nm. The ICCD had mination R2 was about 0.95 using only 3 images for each crank
an array size of 10241024 pixels. This optical assessment allowed angle. While using 10 images the R2 was equal to 0.99.
a spatial resolution around 70 mm/pixel. The ICCD spectral range
spread from UV (180 nm) up to visible (700 nm). The same ICCD 2.3. Exhaust measurement systems
camera was used for 2D ame visualization. In order to evaluate the
ame front speed, a retrieving procedure was set up to collect the Steady-state measurements of CO, CO2, HC, O2 (oxygen) and
optical data. For all the optical measurements, the synchronization NOx(nitrogen oxides) were performed at the raw exhaust by
between the camera and the engine was made by a crank angle gaseous analyzers (Table 2). CO, CO2 and HC were measured by
encoder signal through a delay unit. The camera was not cycle NDIR (non-dispersive infrared detectors). NOx and O2 were
resolved so each image was detected at a xed crank angle of measured by means of an electrochemical sensor. An opacimeter
different engine cycles. was used to measure the opacity of the exhaust gas; the opacity
A post-processing of the optical data was carried out using an percentage was converted in particulate mass concentration using
algorithm properly developed in Labview to obtain geometrical empirical relations [20].
information from each image, such as the mean radius of ame
front. As regards the combustion evolution, every image was pro-
2.4. Fuels
cessed subtracting the background and xing a threshold of 5%
with respect to its maximum light intensity. The background was
The engine was fueled with four different fuels: gasoline,
detected in motored condition before the acquisition of all the se-
ethanol, methane and a blend of 20% v/v of hydrogen in
quences. This methodology allows to evaluate only the 2D projec-
tion of the actual ame front [19] and therefore allows to obtain the Table 2
ame area and the ame outline. The ame radius was obtained for Exhaust meters specications.
each image by the average on six directions. The ame radius curve Pollutant Measurement technique Measurement range Resolution
was calculated by a non-linear regression analysis (third degree
CO Non-dispersive infrared 0e10% 0.01%
polynomial function) using the ame radius extracted from the CO2 Non-dispersive infrared 0e20% 0.1%
average image for each crank angle. The ame front propagation HC Non-dispersive infrared 0e20,000 ppm 1 ppm
speed was calculated as the time rate of change of the ame radius. O2 Electrochemical 4e22% 0.1%
Considering that each single frame belongs to a different engine NOx Electrochemical 0e9000 ppm 1 ppm
Opacity Photodiode detector 0e100% 0.1%
cycle, the cycle-to-cycle variability was taken into account; a
104 F. Catapano et al. / Energy 88 (2015) 101e110

methane, M20H2. The main properties of the fuels are listed in lubrication of the piston, are mainly evident during methane
Table 3: combustion.
The ame kernel in case of a good mixing of the charge moves
from the spark plug with a radial like behavior. However, for all the
2.5. Operating conditions fuels the images highlight that the ame growth is asymmetric
with deviations of the ame's shape from a circle. Similar results
The experimental investigations were carried out at 2000 rpm are also observed in Ref. [22]. The boundaries of the ame appear
at low and full load. For all the test cases, EOI (the end of injection) wrinkled. Looking at the images it is evident that the combustion
and SOS (start of spark), 250 cad (crank angle degree) BTDC (before evolution is strongly inuenced by the fuels. In particular, the
top dead center) and 24 cad BTDC, respectively, were kept constant distortion of the ame boundaries is mainly observed for methane.
for all the fuels in order to better analyze the effect of the fuel on the Moreover, wrinkling boundaries of the ame are evident mainly for
combustion process. The fuel injection occurred in the intake gasoline and methane. The distortion from a circle can be due to
manifold at xed pressure of 3.5 bar for gasoline and ethanol, and at several parameters. Taking in to account that all the fuels have
1.5 bar for methane and M20H2. The DOI (duration of injection) was enough time for the mixing with the air and the evaporation,
properly chosen to obtain a stoichiometric equivalence ratio, as because of the port fuel injection, this asymmetry can be mainly
measured by a lambda sensor installed at the engine exhaust. For ascribed to the turbulence due to the piston motion, which affects
ethanol, the DOI was longer with respect to gasoline because of its the mixture motion eld. This effect is stronger for methane since
lower stoichiometric airefuel ratio: 9 vs 14.7 (Table 3). However, ame growth is quite slower with respect to the other fuels. The
the amount of energy per kg of stoichiometric mixture is similar for more wrinkled boundaries of the ames of methane and gasoline
ethanol and gasoline because of the smaller Lower Heating Value of can be explained taking into account the ame thickness [23] that
ethanol with respect to gasoline (about 26.7 MJ/kg vs 42 MJ/kg). is higher for ethanol and M20H2, and therefore the ame is less
The DOI was longer also for gaseous fuels because of the lower sensitive to the local turbulence and consequently less irregular.
density. The analysis of the ame front radius can give useful informa-
tion about the early combustion phase that is not well resolved by
thermodynamic analysis. The comparison between the ame front
3. Experimental results
radius and the ame front propagation speed for gasoline, ethanol,
methane and M20H2 is shown in Fig. 4.
3.1. Flame front propagation speed
The ame front propagation speed shows a similar behavior for
all the investigated fuels. In particular, it sharply rises, reaches a
The effect of the fuels on the combustion process was studied by
maximum and then decreases because of the increased pressure of
means of optical measurements that allow a high spatial and
the unburned mixture, which slows down the ame front propa-
temporal resolution characterization. The ame front propagation
gation when it is near to the cylinder wall. The ame front propa-
speed, an important parameter in the spark ignition engines, was
gation depends mainly on turbulence, air-fuel ratio (l) and fuel
evaluated from the analysis of the digital images.
properties; in particular, it increases with the increase of turbu-
A selection of natural ame images at typical crank angles for
lence and the decrease of l [24]. At 2000 rpm low load, the tur-
gasoline, ethanol, methane and M20H2 at 2000 rpm full load is
bulence in the intake air is not so strong. For all the fuels the l value
shown in Fig. 3.
should be close to stoichiometric all over the combustion chamber.
For all the fuels, the luminous arc represents the evidence of the
The long time elapsed between the injection and the ignition (SOS)
spark ignition. It occurs around 24 cad BTDC (SOS) and it persists
should allow the formation of a homogeneous mixture. However,
until 8 cad ASOS (after SOS) when the ame kernel is clearly visible,
the presence of rich zone is evidenced by the presence of luminous
even if its luminosity is much lower than the spark. For gasoline
ame (Fig. 3). The ame speed depends also on the fuel properties.
fueling are evident luminous ames, known as pool ames, close to
It is evident that the ame radius propagation is much slower for
the intake valves due to the diffusive combustion of the deposited
methane with respect to the other fuels; this result depends on the
liquid fuel [21]. High intensity bright spots, due probably to the
different laminar burning velocity. The laminar velocity is, in fact, a
bronze of the Teon-bronze rings used in the optical engine for the
property of the fuels. It was dened as the speed at which an un-
stretched laminar ame will propagate through a quiescent
Table 3 mixture of unburned reactants. The ame of the M20H2 propagates
Fuel properties. faster than methane highlighting that the hydrogen addition ac-
celerates the combustion process. This can be due to the higher
Fuel property Gasoline Ethanol Methane Hydrogen
laminar ame speed of hydrogen with respect to methane (Table 1).
Formula C4 a C12 C2H5OH CH4 H2
Moreover, the addition of hydrogen stimulates the formation of H*,
Molecular weight [kg/kgmol] 100e105 46.07 16.04 2.02
Carbon [mass %] 85e88 52.2 75 e O* and OH* radicals. These radicals are effective on accelerating the
Hydrogen [mass %] 12e15 13.1 25 100 chain reactions and consequently improving the local laminar
Oxygen [mass %] 2.7 34.7 e e ame speed [24]. The ame propagation speed of M20H2 is higher
Density [kg/Stm3] 720e775 790 0.67 0.08 also with respect to the other fuels. This result could be due to the
Boiling point [ C] 27e225 78 161.4 252.9
Vapour pressure. [kPa at 38  C] 48e103 15.9 e e
different ammability limits of the fuels. The ammability limits of
Specic heat [kJ/kgK] 2 2.4 2.22 1.43 hydrogen are, in fact, wider with respect to the other fuels and at
Viscosity [mPa$s at 20  C] 0.37e0.44 1.19 0.011 0.009 the same time the ignition energy is lower (Table 1). Moreover, the
Lower heating value [MJ/kg] 42e44 26.7 50 120 enhanced laminar ame speed due to hydrogen addition could
Auto ignition temperature [ C] 257 423 600 645
further improve the local reaction rate [25]. For methane, the ame
Research octane number 98 108.6 >120 >130
Flammability limit [vol %] 1.4/7.6 4.3/19 5.3/14 4/75 propagation speed is lower with respect to the other fuels at both
Minimum ignition energies [mJ] 0.24 0.22 0.28 0.02 the engine operating conditions. On the other hand, a different
Energy density [MJ/l] 32.4 20.9 0.036 5.6 behavior is observed for gasoline and ethanol with the engine
Stoichiometric air/fuel 14.7 9 17.24 34.20 operating conditions. In particular, at low load, the propagation
Laminar burning velocity in air [cm/s] 35 41 37 250
speed of gasoline and ethanol ame is lower with respect to
F. Catapano et al. / Energy 88 (2015) 101e110 105

Fig. 3. Selection of combustion images for gasoline, ethanol, methane and a M20H2 at full load.
106 F. Catapano et al. / Energy 88 (2015) 101e110

Fig. 4. Front ame diameter (left side) and velocity (right side) for gasoline, ethanol, methane and M20H2 at low load (upper) and full load (lower).

M20H2. This result can be due to the inhomogeneity of the mixture. left side shows the natural luminosity emitted from gasoline
This effect is more evident for ethanol because the typical low in- combustion in three typical locations of the combustion chamber at
cylinder temperature of this condition is further reduced by the the same time; on the right side is shown a comparison between
cooling effect of ethanol caused by its higher heat of vaporization the chemiluminescence emission spectra of all the tested fuels
compared to other fuels. It is worth noting that because of its lower detected at the same location and time.
stoichiometric value, a larger quantity of ethanol is injected. This The spectra show two peaks centered at 280 nm and 309 nm due to
means that ethanol vaporization cools the air, more mass is drawn OH* emission. A broadband emission, from 250 to 500 nm, was
into the cylinder, so increasing the volumetric efciency. But at the detected for gasoline, methane and M20H2. This band was due to the
same time, the vaporization of the fuel is worsened creating in- convolution of HCO Vayda bands from 250 nm to 410 nm, and HCHO
homogeneity in the mixture that slows down the ame front Emeleus bands from 340 nm to 523 nm [27]. Moreover, CH* radical
propagation. At 2000 rpm full load, the higher turbulence intensity emission was evidenced by the peak centered at 431 nm. For the
and temperature with respect to the previous condition reduces the ethanol, the HCO bands as well as CH* are absent. The spatial evolution
differences between the ame front propagation and speed of of the radical emission was shown in Fig. 5-left side. The emission
gasoline, ethanol and M20H2. This is probably due to a better intensity is strongly related to the combustion chamber location. In
evaporation and mixing between the fuel and the air, which leads particular, it is higher at the spark plug and lower at the intake valves.
to a more homogeneous mixture that reduces the inuence of the This behavior was observed for all the investigated fuels. It can be
fuel chemical properties. ascribed to the asymmetric evolution of the ame, which is moved by
the air motion towards the exhaust valves, as observed in the com-
3.2. Flame emission spectroscopy bustion images. From the comparison of the emission spectra of the
investigated fuels, Fig. 5-right side, it is evident the relationship be-
The natural emission spectroscopy in the UVevisible range was tween the radical OH* and the fuel. The emission intensity is higher for
applied in order to characterize the combustion process. This diag- M20H2 over all the wavelength range conrming the faster kernel
nostic technique allows, in fact, an in-depth understanding of the evolution of the fuel. The lower chemiluminescence intensity is
combustion process by means of the detection and the monitoring of observed for the ethanol. The lower OH* emission intensity of ethanol
the radical species produced during the combustion process. can be ascribed to the lower in-cylinder temperature. On the other
Particular attention was paid to the chemiluminescence emission hand, the higher OH* emission for gaseous fuels can be due to the
from OH* and CH*, which are considered the more signicant radical better mixing and efcient combustion. It is worth noting that the OH*
species in the combustion process and emission formation [26]. radical emission increases signicantly adding hydrogen to the
Fig. 5 shows the chemiluminescence emission spectra detected methane, conrming the larger radical pools that improve the local
during the ame front propagation. In particular, the picture on the laminar ame speed. In terms of CH* emissions, the highest values are
F. Catapano et al. / Energy 88 (2015) 101e110 107

Fig. 5. Natural emissions spectra in three locations of the combustion chamber for gasoline at the same time (left side), and for gasoline, ethanol, methane and M20H2 at the same
time and location (right side).

measured for gasoline both because of the different carbon contents in 3.4. Exhaust emissions
the fuel and the worst break of the fuel into simple hydrocarbons.
Fig. 7 shows the effect of different fuels on the emissions
measured at the engine exhaust.
3.3. Fuel consumption For both engine operating conditions, the CO concentration of
the alternative fuels is lower with respect to that of gasoline. This
The BSFC (brake specic fuel consumption) measured for all the can be ascribed to their more complete combustion due to the
tested fuels and operating conditions is shown in Fig. 6. larger oxygen content for the ethanol and the better homogeniza-
For both the tested conditions, the fuel consumption is higher tion with respect to gasoline of the gaseous fuels.
for the ethanol with respect to the other fuels because of its lower CO2 emissions are directly related to the combustion efciency,
stoichiometric air/fuel ratio. Particularly, the ethanol stoichiometric the BSFC and the quantity of carbon present in the fuel. The lower
air/fuel ratio is smaller than gasoline, 9 vs. 14.6; at the same time, it CO2 emission for gaseous fuels can be ascribed to the lower BSFC as
has a smaller LHV, about 26.8 MJ/kg vs. 43.2 MJ/kg of gasoline. It well as to the lower carbon content with respect to gasoline. For
was evaluated that the amount of energy per kg of stoichiometric ethanol, the lower carbon content is partially compensated by the
mixture is similar for ethanol and gasoline, 2.68 MJ/kg vs. 2.77 MJ/ larger fuel consumption. For this reason, the CO2 emissions can be
kg, respectively. mainly due to the combustion efciency. In particular, the lower
Valuable information can be obtained also from the analysis of combustion efciency at low load, mainly due to the poor ethanol
the EC (energy consumption) dened as EC BSFC * LHV, shown in evaporation at low temperature, provides a lower CO2 emission
Fig. 6. At low load, the ethanol has higher EC than gasoline indi- with respect to gasoline. On the other hand, at full load the CO2
cating a less efcient combustion because of the inhomogeneity of emissions are larger for ethanol. In this case, the easier evaporation
the mixture, evidenced also by the optical data. The EC for M20H2 is of ethanol leads to the formation of a more homogeneous mixture.
lower with respect to the gasoline because of the more efcient Moreover, its larger oxygen content contributes to a more complete
combustion due to a uniform mixture formation and fast ame combustion.
propagation. On the other hand, for methane, the slow ame The HC emissions are larger in both conditions for the gasoline
propagation compensates the benecial effect of the good mixing because of the poor mixing that generates more locally rich regions.
typical of the gaseous fuel so providing an EC quite comparable to The HC concentration is lower for ethanol fueling at all the inves-
that of gasoline. At full load, the higher turbulence intensity and tigated engine conditions. The larger oxygen content in ethanol
temperature with respect to the previous condition reduces the increases the local oxygen content also in the fuel-rich regions and
differences between the EC of gasoline and ethanol. However, the consequently the combustion is more complete [28,29]. Moreover,
M20H2 combustion is also in this case the most efcient thanks to the ethanol evaporates more easily producing a more homogenous
the hydrogen properties.

Fig. 6. BSFC (left side) and EC (right side) for gasoline, ethanol, methane and M20H2.
108 F. Catapano et al. / Energy 88 (2015) 101e110

Fig. 7. Exhaust emissions for gasoline, ethanol, methane and M20H2.

mixture. The gaseous fuels show larger HC emission with respect to provided by the absence of low volatile compounds and the larger
ethanol because of the presence of the methanic HC at the exhaust. oxygen content. The combustion is enhanced also in the rich re-
For both the conditions, as expected, NOx emissions have the gions of the combustion chamber, such as close to the intake valves,
opposite behavior with respect to CO emissions. Regarding the fuel where the impingement of the fuel causes the well-known pool re
effect, it has to be taken into account the oxygen content and the [30e32].
total low heating value of the fuel related to the temperature
reached in the cylinder. As known, the oxygen content boosts the
3.5. Fuel ranking
NOx formation. For this reason for the ethanol, NOx emissions are
larger with respect to the gasoline. On the contrary, less NOx are
In order to better analyze the effect of the alternative fuels on
formed during methane combustion. The slower combustion speed
the combustion process a classication of the fuels was performed.
and the lower total heating value of the charge, due to the fuel
In particular, for each parameter, such as BSFC, EC and the exhaust
chemical properties, cause a lower combustion temperature. In
emissions, a relative weight factor was arbitrarily attributed
both conditions, the effect of hydrogen addition is evident. In
(Table 4). Particular attention was paid to the pollutant emissions.
particular, the NOx emissions are larger for M20H2 because of the
faster combustion speed and the improved efciency-gives high
Table 4
local temperature. Relative weight factor for fuel ranking.
The PM (particulate matter) emissions were evaluated from the
Parameter Coefcient
smoke opacity. It is interesting to observe that the values are very
low for all the fuels in both the operating conditions. For gasoline IMEP 2
BSFC 2
fuel, the opacity and optical data show a quite good agreement.
EC 2
However, for the gasoline the PM emissions are larger than other NOx 3
fuels. This can be ascribed to the presence of fuel rich zones caused CO 3
by the impingement of the fuel close to the intake valves that burn CO2 1
in rich condition producing diffusive ames. For the ethanol, the PM 3
HC 3
lower PM emissions are due to the better vaporization of the fuel
F. Catapano et al. / Energy 88 (2015) 101e110 109

hydrogen in M20H2 produces higher temperature and then in


larger NOx emissions with respect to the other investigated fuels.
A classication of fuels, which takes into account the perfor-
mance and the exhaust emissions, highlighted that M20H2 has the
best behavior in terms of both energy sustainability and pollutant
emissions reduction. In particular, it keeps the typical low emis-
sions of methane fuel and takes advantage of the hydrogen addition
that improves the combustion efciency.

Acknowledgments

Authors are grateful to Mr. Carlo Rossi and Mr. Bruno Sgammato
for the engine assessment and for the support in the experimental
activity.
Fig. 8. Fuels ranking.

Abbreviations
Then, a score between 1, for the worst, and 4, for the best, was
attributed for each parameter. As example, for NOx emissions at full AFR air-fuel ratio
load, the score 4 was given to the methane that shows the lowest ASOS after start of spark
NOx emissions; on the other hand, the score 1 was attributed to BSFC brake specic fuel consumption
M20H2 that has the largest NOx emissions. Therefore, the weighted BTDC before top dead center
average of the score, where the weight factors are the coefcients cad crank angle degree
shown in Table 4, was calculated and the results are shown in Fig. 8. CDM crank angle degree marker
This analysis highlights that M20H2 is the best fuel: it has the CO carbon monoxide
same advantages of methane, low emissions and fuel consum CO2 carbon dioxide
ption, saving the performance thanks to the more efcient CoV coefcient of variance
combustion. DOI duration of injection
EC energy consumption
EOI end of injection
4. Summary/conclusion HC hydrocarbons
ICCD intensied charge couple device
The aim of the paper is the analysis of the effect of the alter- IMEP indicated mean effective pressure
native fuels on the engine performance, the pollutant emissions LHV low heating value
and the energy sustainability. The investigation was carried out on M20H2 Blend of 20% v/v hydrogen in methane
an optically accessible small single-cylinder, four-stroke engine. It NDIR non-dispersive infrared detectors
was equipped with the cylinder head of a PFI commercial 250 cc NOx nitrogen oxides
engine. It was fueled with gasoline, ethanol, methane and a blend of PFI port fuel injection
hydrogen at 20% v/v in methane. Tests were carried out at PM particulate matter
2000 rpm at low and full load. SI spark ignition
A detailed analysis of the effect of the fuels on the combustion SOS start of spark
process was carried out by means of the ame emission spectros- TDC top dead center
copy in the UVevisible range and 2D-digital imaging measure-
ments. The gaseous and particulate emissions were measured at
the exhaust by means of conventional instruments. References
The ethanol ame front propagation speed is quite similar to
that of the gasoline. The methane is characterized by the slowest [1] Pourkhesalian AM, Shamekhi AH, Salimi F. Alternative fuel and gasoline in an
SI engine: a comparative study of performance and emissions characteristics.
ame speed; higher values were measured by adding 20% v/v of Fuel 2010;89:1056e63.
hydrogen, which accelerates the combustion process. Moreover, [2] Balki MK, Sayin C. The effect of compression ratio on the performance,
spectroscopic measurements have highlighted that the OH* radical emissions and combustion of an SI (spark ignition) engine fueled with pure
ethanol, methanol and unleaded gasoline. Energy 2014;71:194e201.
emission increases signicantly by adding hydrogen in the [3] Ozsezen AN, Canakci M. Performance and combustion characteristics of
methane fuel, contributing to the improvement of the local laminar alcohol-gasoline blends at wide-open throttle. Energy 2011;36:2747e52.
ame speed. [4] Gong C, Huang K, Jia JL, Su Y, et al. Regulated emissions from a direct-injection
spark-ignition methanol engine. Energy 2011;36:3379e87.
Energy consumption data showed the more efcient combus- [5] Cooney CP, Worm YJ, Naber JD. Combustion characterization in an internal
tion of the ethanol with respect to gasoline because of its larger combustion engine with ethanolegasoline blended fuels varying compression
oxygen content and the easier evaporation mainly due to the lower ratios and ignition timing. Energy Fuels 2009;23(5):2319e24.
mez Gonzalez U. Combustion and emissions
[6] Schifter L, Diaz R, Rodriguez JP, Go
boiling point of ethanol. Meanwhile, the lower energy consumption behavior for ethanolegasoline blends in a single cylinder engine. Fuel
for the gaseous fuel can be ascribed to the better fuel-air mixing 2011;90:3586e92.
giving a more complete combustion; this effect is enhanced for [7] Celik MB. Experimental determination of suitable ethanolegasoline blend rate
at high compression ratio for gasoline engine. Appl Therm Eng 2008;28:
M20H2 by the faster combustion. The lower CO and CO2 emissions
396e404.
also evidence the more efcient combustion of the alternative fuels. [8] Hsieh WD, Chen RH, Wu TL, Lin TH. Engine performance and pollutant
On the other hand, NOx emissions were larger for ethanol with emission of an SI engine using ethanolegasoline blended fuels. Atmos Environ
respect to gasoline because of the larger oxygen content. A NOx 2002;36:403e10.
[9] Karim GA, Wierzba I. Comparative studies of methane and propane as fuels for
emissions reduction was observed, instead, for methane fueling spark ignition and compression ignition engines. SAE Trans 1983;92. SAE
because of the low combustion temperature. The presence of Paper 831196 (SP-548).
110 F. Catapano et al. / Energy 88 (2015) 101e110

[10] Rousseau S, Lemoult B, Tazerout M. Combustion characteristics of natural gas [21] Zhao F, Lai MC, Harrington DL. Automotive spark-ignition direct-injection
in a lean burn spark-ignition engine. Proc Institution Mech Eng Part D J gasoline engines. Prog Energy Combust Sci 1999;25:437e562.
Automob Eng 1999;213(D5):481e9. [22] Aleiferis PG, Rosati MF. Flame chemiluminescence and OH LIF imaging in a
[11] Roser R., Hydrogen-enriched natural gas bus demonstration, NRG Technolo- hydrogen-fueled spark-ignition engine. Int J Hydrog Energy 2012;37:
gies, Inc. 681 Edison Way Reno, NV 89502. 1797e812.
[12] Dimopoulos P, Boulouchos K, Rechsteiner C, Soltic P, et al. Combustion char- [23] Miao H, Jiao Q, Huang Z, Jiang D. Effect of initial pressure on laminar com-
acteristics of hydrogen-natural gas mixtures in passenger car engines. SAE bustion characteristics of hydrogen enriched natural gas. Int J Hydrog Energy
Technical Paper 2007-24-0065. 2007. http://dx.doi.org/10.4271/2007-24- 2008;33:3876e85.
0065. [24] Heywood JB. Internal combustion engine fundamentals. New York: McGraw-
[13] Andersson T. Hydrogen addition for improved lean burn capability on natural Hill; 1988.
gas engine. Rapport SGC 134. ISSN1102-7371. Lund Institute of Technology; [25] Emadi M, Karkow D, Salameh T, Gohil A, Ratner A. Flame structure changes
2002. resulting from hydrogen-enrichment and pressurization for low-swirl pre-
[14] Park C, Kim C, Choi Y, Won S, Moriyoshi Y. The inuences of hydrogen on the mixed methane-air ames. Int J Hydrog Energy 2012;37:10397e404.
performance and emission characteristics of a heavy duty natural gas engine. [26] Matsui Y, Kamimoto T, Matsuoka S. Formation and oxidation processes of 1
Int J Hydrog Energy 2011;36:3739e45. soot particulate in a DI diesel enginedan experimental study via the two-
[15] Xu J., Zhang X., Liu J., Fan L., Experimental study of a single-cylinder engine color method. SAE Transaction SAE Paper 820464. 1982. p. 285e98.
fueled with natural gas-hydrogen mixtures, Int J Hydrog Energy 201; 35: [27] Gaydon AG. The spectroscopy of ames. Chapman and Hall ltd; 1957.
2909e2914. [28] Naja GB, Ghobadian T, Tavakoli DR, Buttsworth Yusaf TF, Faizollahnejad M.
[16] Ma F, Wang M, Jiang L, Deng J, Chen R, Naeve N, et al. Performance and Performance and exhaust emissions of a gasoline engine with ethanol
emission characteristics of a turbocharged spark-ignition hydrogen enriched blended gasoline fuels using articial neural network. Appl Energy 2008;86:
compressed natural gas engine under wide open throttle operating condi- 630e9.
tions. Int J Hydrogen Energy 2010;35:12502e9. [29] Bayraktar H. Experimental and theoretical investigation of using gaso-
[17] Wang J, Chen H, Liu B, Huang Z. Study of cycle-by-cycle variations of a spark lineeethanol blends in spark-ignition engines. Renew Energy 2005;30:
ignition engine fueled with natural gas-hydrogen blends. Int J Hydrog Energy 1733e47.
2008;33:4876e83. [30] Huang Y, Matthews R, Ellzey J, Dai W. Effects of fuel volatility, load, and speed
[18] Wang S, Ji C, Zhang J, Zhang B. Comparison of the performance of a spark- on HC emissions due to piston wetting. SAE Technical Paper; 2001. n. 2001-
ignited gasoline engine blended with hydrogen and hydrogen-oxygen mix- 01-2024.
tures. Energy 2011;36:5832e7. [31] Stanglmaier RH, Roberts CE, Moses CA. Vaporization of individual fuel drops
[19] Sementa P, Vaglieco BM, Catapano F. Thermodynamic and optical character- on a heated surface: a study of fuel-wall interactions within direct-injected
izations of a high performance GDI engine operating in homogeneous and gasoline (DIG) engines. SAE Technical Paper; 2002. p. 185e94. n.2002-01-
stratied charge mixture conditions fueled with gasoline and bio-ethanol. 0838.
Fuels 2012;5(96):204e19. [32] Urban BD, Kroenlein K, Kazakov A, Dryer FL, et al. Sooting behavior of ethanol
[20] Mo rsch O, Sorsche P. Investigation of alternative methods to determine par- droplet combustion at elevated pressures. Microgravity Sci Technol 2004;XV/3.
ticulate mass emissions. UNECE/WP.29/GRPE report 2001. 2001.

You might also like