Applied Energy 136 (2014) 4350

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Lean burn performance of a hydrogen-blended gasoline engine

at the wide open throttle condition
Shuofeng Wang, Changwei Ji , Bo Zhang, Xiaolong Liu
College of Environmental and Energy Engineering and Key Laboratory of Beijing on Regional Air Pollution Control, Beijing University of Technology, Beijing 100124, China

h i g h l i g h t s
 The performance of a H2-blended gasoline engine at the WOT condition was studied.
 The engine became run stable after the H2 addition.
 H2 addition resulted in the raised thermal efciency for the gasoline engine.
 Both number and mass of particulate emissions were reduced by the H2 addition.
 H2 addition reduced the engine knocking tendency at full loads.

a r t i c l e

i n f o

Article history:
Received 21 March 2014
Received in revised form 11 August 2014
Accepted 14 September 2014
Available online 29 September 2014
Keywords:
Hydrogen
Gasoline
Lean burning
Emissions
Wide open throttle

a b s t r a c t
The performance of a hydrogen-blended gasoline engine at lean and the wide open throttle conditions
was investigated. A hydrogen port-injection system was adopted to introduce the hydrogen into each
cylinder. The engine was operated at 1400 rpm and two hydrogen blending levels of 0% and 3%. The
excess air ratio was raised from 1.00 to about 1.45 for a given hydrogen addition fraction. The test results
demonstrated that the hydrogen blending contributed to the raised thermal efciency and shortened
ame development and propagation durations. An increased brake mean effective pressure was found
after the hydrogen addition only at lean conditions. For both stoichiometric and lean conditions, the
hydrogen blending was benecial for reducing the engine cyclic variation. This provides a possibility
to run a hydrogen-blended gasoline engine with the fully opened throttle position and control the engine
torque only by adjusting the excess air ratio. Toxic emissions including HC, CO and particulate were
reduced after the hydrogen blending.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
The daily increased energy crisis and demand on environmental
protection have stimulated studies on clean and renewable alternative fuels for vehicles. In recent years, many efforts have been put
on developing the alternative fuel for internal combustion engines,
such as ethanolgasoline engines [13], methanol engines [4,5],
and dimethyl ether-blended engines [6]. Among all fuel candidates,
hydrogen is generally believed to be a promising alternative fuel for
the internal combustion engines. Different from fossil fuels,
hydrogen can be produced through kinds of ways [710], such as
water electrolysis, fuel reforming and biological hydrogen production. This makes the hydrogen-based power system become important to the nation energy safety. There are three typical ways for
Corresponding author. Tel./fax: +86 1067392126.
E-mail address: chwji@bjut.edu.cn (C. Ji).
http://dx.doi.org/10.1016/j.apenergy.2014.09.042
0306-2619/ 2014 Elsevier Ltd. All rights reserved.

applying hydrogen on vehicles [1114], which are fuel cells, pure

hydrogen engines and hydrogen-blended engines. Generally, the
fuel cell-powered vehicles produce few harmful emissions during
the driving time. However, the high price and limited life span of
fuel cells are generally barriers for its commercialization at present.
Besides, the limited hydrogen infrastructure distribution also
makes the relling of fuel cell vehicles difcult in many countries.
Furthermore, spark ignition engines are still used worldwide. This
makes improving the thermal efciency and reducing toxic emissions of internal combustion engines more important for ensuring
the nation energy safety and environmental protection. The combustion of pure hydrogen engines is cleaner than the traditional
gasoline and diesel engines, but the pure hydrogen engines always
suffer the high NOx emissions and dropped power output [15].
Besides, abnormal combustion phenomenon, such as pre-ignition
and knock, is prone to be appeared in pure hydrogen engines
[16,17]. Compared with fuel cells and pure hydrogen engines, the

44

S. Wang et al. / Applied Energy 136 (2014) 4350

hydrogen-blended engine is also capable of gaining less toxic emissions and better fuel economy [1820]. Generally, converting a conventional gasoline engine to be fueled with hydrogengasoline
blends requires adding a hydrogen injection system to the engine,
and the hydrogen could be stored in the hydrogen cylinders or
provided by the onboard hydrogen producer, which is usually introduced to the hydrogen injectors through a stainless steel made
hydrogen supply system. Besides, to accomplish the hydrogen
injection, the program of engine electronic control unit has to be
improved accordingly. Although it needs time and work to accomplish these work, as it does not need to redesign the engine or massive improvements in the powertrain system, realizing the
hydrogen blending to a conventional gasoline engine is feasible
and not a very tough work. Thus, the hydrogen blending is a promising approach to better engine performance in the near future.
Basically, improving the combustion process of fossil fuel is
important on enhancing the engine overall performance. The effect
of hydrogen addition on the combustion properties of fossil fuels
has been done by Huang et al. [21,22]. It was found that the hydrogen blending availed accelerating the ame propagation for natural
gas-air mixtures. Besides, a combustion simulation model was built
and calibrated. Based on this model, they observed that the hydrogen blending to methane contributed to promoting the formation of
OH, O, and H in the ame. These could help shorten the combustion
duration of fuelair mixtures and therefore reducing the engine
cooling and exhaust losses. Furthermore, they also found that the
ame shape and position became stable after the hydrogen enrichment. This is generally helpful for easing the engine cyclic variation.
Except for the investigations on basic combustion properties of
hydrogen-blended fossil fuels, there are also some investigations
which have shown the effect of hydrogen enrichment on engine performance. Regarding the daily increased urgency for energy saving,
it was demonstrated that the addition of hydrogen was capable of
enhancing thermal efciency of the natural gas, LPG and gasoline
engines [2328]. According to Lata et al. [2427], a 17% improvement in thermal efciency was found for the LPG engine after the
hydrogen addition. Reasons for the improved engine thermal efciency were generally believed to the shortened combustion duration and enhanced combustion completeness which contributed
to the dropped cooling and exhaust losses. Furthermore, because
of the short quenching distance of hydrogen, the addition of hydrogen was also proved to be helpful for reducing the engine HC and CO
emissions [29,30]. Navarro et al. [31] also found that because the
hydrogen was a carbonless fuel, the addition of hydrogen also
availed reducing carbon dioxide emission from the engine. However, because of the high ame temperature of hydrogen, NOx emissions from the hydrogen-enriched fossil fuel engines were generally
higher than those from conventional engines [29,30]. Fortunately,
because of the wide ammability of hydrogen enabled the hydrogen-enriched engine to run stably at much leaner conditions
[32,33], NOx emissions from the hydrogen-enriched engines could
be controlled by adopting lean combustion strategy. Investigations
done by Diguez et al. [28] found that NOx emissions were
decreased to near zero when the hydrogen-natural gas blendsfueled engine was run at very lean conditions. To further improve
the performance of hydrogen-enriched engines, Ji et al. [34] proposed a control strategy which changes the hydrogen-to-gasoline
ratio according to the engine working conditions for the hydrogen-enriched gasoline engine. By adopting this controlling strategy,
the engine fuel economy was improved and the toxic emissions during the legislated new European driving cycles (NEDC) were obviously reduced for the hydrogen-blended gasoline engine-powered
vehicle.
Operating an engine at the wide open throttle (WOT) condition
is very effective for reducing the engine pumping loss [35], especially for the naturally-aspirated engines. Moreover, the enhanced

charge motion at the WOT condition also leads to the dropped

residual gas fraction which contributes to the improved engine
combustion. According to previous study [36], because of the
increased charge heating capacity, the application of exhaust gas
recirculation (EGR) was helpful for inhibit the knocking of spark
ignition engines. Xie et al. [37] tried to run a methanol engine at
the WOT condition and control the engine load through adjusting
the spark timing and the ratio of EGR. They conrmed that the
engine brake mean effective pressure can be changed between
0.36 and 0.96 MPa at the un-throttled condition by properly controlling the spark timing and EGR ratio. However, according to
investigations carried out by Fontana and Galloni [35], the engine
volumetric efciency tended to be decreased by increasing EGR
ratio at the WOT condition. This would lead to higher residual
gas fractions which may deteriorate in-cylinder combustion. Generally, the challenge for running an SI engine at the WOT condition
is knocking. Due to the elevated temperature in combustion chamber, the end gas tends to be auto-ignited. Particularly, for the
hydrogen engines, the high ame temperature of hydrogen may
make the pure hydrogen engines are much easier to encounter
knock at high loads [38].
However, although running an SI engine at the WOT condition
is benecial for enhancing the combustion and reducing the pumping loss, there are limited papers examining the operating characteristics of an engine fueled with hydrogengasoline mixtures
under the WOT condition. As running an engine at the un-throttled
condition is effective on promoting the engine fuel economy and
reducing the emissions, this paper experimentally investigated
the operating characteristics of an engine fueled with hydrogen
gasoline mixtures under the WOT condition. As the application of
lean combustion contributes to the reduced cylinder temperature
[39] which may be helpful for overcoming knocking in spark ignition engines, the tests are mainly conducted under lean conditions.

2. Experimental set-up and methodology

2.1. Experimental set-up
The test engine is a naturally-aspirated Beijing Hyundai manufactured gasoline engine with a displacement volume of 1.6 L, a
rated torque of 143.28 N m at 4500 rpm and a rated power of
82.32 kW at 6000 rpm. Before this experiment, a hydrogen portinjection system was added so that the engine can be fuelled with
the hydrogengasoline blends. The hydrogen injection system contains a hydrogen rail and four hydrogen injectors placed on the
runners of each cylinder. Besides, in order to govern the hydrogen
and gasoline injection durations and spark timing, a hybrid electronic control unit (HECU) was applied. Through changing programs in the HECU, the hydrogen blending level, engine spark
timing, and excess air ratio of the fuelair mixtures could be
adjusted in real time.
A GW160 eddy current dynamometer was applied to govern the
engine speed by putting loads on engines (measurement uncertainty: 0.28 N m and 1 rpm). Two thermal ow meters typed
20N060 and D07-19BM were applied to record the hydrogen and
air ow rates, respectively (measurement uncertainty: <0.1 and
0.02 L/min for air and hydrogen, respectively). The mass ow rate
of gasoline was detected by a FC2210 fuel mass ow meter (measurement uncertainty: <0.33 g/min). The engine in-cylinder pressure was recorded through a Kistler 6117BFD17 piezoelectric
pressure sensor (measurement uncertainty: <0.3 bar). The crankshaft position was recorded by a Kistler 2613B optical encoder
(measurement deviation: <0.01 CA, crank angle resolution:
0.2 CA). Then, to conduct the insight combustion analysis, the
engine crank angle signals and cylinder pressure were sent to a

S. Wang et al. / Applied Energy 136 (2014) 4350

45

DEWE-800 combustion analyzer through screened cables. NOx, CO,

and HC were detected by a Horiba MEXA 7100 DEGR emission analyzer (sensitivity: 1 ppm, measurement uncertainty: <1%). As the
particulate emissions from spark ignition engines are also harmful
to the environment and human healthy, the particulate emissions
were measured through a DMS 500 particulate analyzer.
2.2. Methodology
The test was carried out when the warming up procedure was
complete. The basic fuel used in this study is the commercial gasoline produced by Sinopec, which has a research octane number of
92. Table 1 lists the properties of hydrogen and gasoline [40]. The
engine was operated under 1400 rpm and the WOT condition. The
hydrogen injection duration was changed to achieve two hydrogen
volume fractions in the total intake (aH2) of 0% and 3%. For each
aH2, the ow rate of gasoline was adjusted so that the global excess
air ratio (k) of gasolinehydrogenair mixtures can be varied from
1.00 to about 1.45. To ensure that the best performance could be
obtained, spark timings were adjusted to be relevant spark timings
for the maximum brake torque (MBT) without knocking. The
initiation of gasoline and hydrogen injections was set to be same
timing as that of the original engine at the WOT condition to minimize the effect of injection timing on the vaporization and mixing
of charge. Since the main throttle was kept at fully opened and the
exhaust was directly expelled to the ambient, both inlet and
exhaust pressures could be seen the same as the ambient pressure
of 1 bar. aH2 and k are dened by the following equations:

aH2 V H2 =V H2 V air   100%

k V air qair =V H2 qH2 AFst;H2 mgas AFst;gas

In Eqs. (1) and (2), the stoichiometric air-to-fuel ratios of

gasoline and hydrogen are represented by AFst,gas and AFst,H2,
respectively; the volume ow rates of air and hydrogen at the normal condition are symbolized by Vair and VH2 [L/min]; the densities
of air and hydrogen at the normal condition are indicated by qair
and qH2 [g/L], and the gasoline mass ow rate is represented by
mgas [g/min].
3. Results and discussion
3.1. Brake mean effective pressure and thermal efciency
The variations of brake mean effective pressure (Bmep) with k
for the engine with 0% and 3% hydrogen blending at 1400 rpm
and the WOT condition are plotted in Fig. 1. It is seen from Fig. 1
that Bmep of the engine blended with 0% and 3% hydrogen are
reduced with the enhancement of k due to the dropped fuel energy
supplement. Besides, when the k is larger than 1.20, Bmep of the
pure gasoline engine shows a sharper decreasing tendency than
that with 3% hydrogen blending. This is because the hydrogen
has a wider ammability than gasoline as it is shown in Table 1.

Fig. 1. Bmep versus k at 1400 rpm and WOT condition.

According to Le Chateliers formula [41,42], blending a fuel with

wide ammability could extend the ammability of basic fuel. As
a result, the gasolinehydrogenair mixtures could be more fully
combusted at high k due to the extended fuel lean burn limit.
Thereby, Bmep of the gasoline engine blended with 3% hydrogen
reduces more gently when k is further elevated from 1.20. Fig. 1
also indicates that Bmep of the gasoline engine blended with 3%
hydrogen is slightly lower than that without hydrogen blending
when k is smaller than 1.20. But when the k is greater than 1.20,
Bmep of the gasoline engine blended with 3% hydrogen is larger
than that of the pure gasoline engine. When k = 1.00, Bmep of
the gasoline engine is decreased by about 1.2% after the hydrogen
blending. Comparatively, when k reaches about 1.45, a 25.1%
enhancement in Bmep is observed with 3% hydrogen blending.
The reason is that the hydrogen has a lower volume energy density. Thus, for a given k, the 3% hydrogen-blended gasoline engine
has a lower fuel energy ow rate (Ef) than the original engine. This
can be seen from Fig. 2 which plots the variations of fuel energy
ow rate with excess air ratio. At the stoichiometric and slightly
lean conditions, both of the gasoline and hydrogengasoline mixtures could be burnt completely. Thus, the hydrogen blending
may result in the dropped Bmep due to the reduced fuel energy
ow rate as it can be found from Fig. 2. However, at high k, the
quickly raised HC emissions which could be found from Fig. 9a
shows that the pure gasoline engines are prone to suffer
incomplete combustion because of the narrow ammability range

Table 1
Key combustion and physicochemical properties of hydrogen and gasoline.
Property

Hydrogen

Gasoline

Molecular weight (g/mol)

Density (kg/m3)
Lower heating value (MJ/kg)
Minimum ignition energy (mJ)
Quenching distance (cm)
Flammability limits in air (vol%)
Unstretched laminar burning velocity (cm/s)
Stoichiometric air-to-fuel ratio (kg/kg)

2.016
0.08
120
0.02
0.06
475
237
34.3

110
720750
44.5
0.24
0.2
1.37.6
3547
14.6

Fig. 2. Ef versus k at 1400 rpm and WOT condition.

46

S. Wang et al. / Applied Energy 136 (2014) 4350

of gasoline. This leads to the sharp decrease in Bmep of the pure

gasoline engine. Comparatively, the reduced HC emissions at high
excess air ratios which is shown in Fig. 9a demonstrates that the
wide ammability of hydrogen permits the gasolinehydrogen
blends to be burned more completely than the pure gasoline.
Therefore, the hydrogen blending avails enhancing Bmep at only
lean conditions.
Fig. 3 depicts the variations of engine brake thermal efciency
(BTE) with k at 1400 rpm and two hydrogen blending levels. It is
seen from Fig. 3 that, for a specied k, BTE is elevated after the
hydrogen blending, particularly at high k. When k = 1.45, BTE is
improved while aH2 rises from 0% to 3%, because the quick combustion of the fuelair mixtures is achieved thanks to the enhanced
ame speed with hydrogen blending. As the ame is propagated
more quickly after the hydrogen addition (see Figs. 4a and 4b),
the fuel burned during the piston downward movement could be
reduced. This contributes to the reduced cooling and exhaust
losses which could be deduced from Figs. 6 and 7. Meanwhile,
the short quenching distance of hydrogen enables the ame to
propagate near the cylinder wall. This also benets the enhanced
combustion completeness at aH2 = 3%. Fig. 3 shows that BTE of
the pure gasoline engine rst increases with k. But when the k is
larger than about 1.20, the further increased k may result in the
dropped BTE for the pure gasoline engine. Comparatively, BTE of
the engine at aH2 = 3% is continuously raised with the increase of
k. This is because the properly increased k helps the fuel to be combusted completely because of the enhanced oxygen fraction. However, at further elevated k, the gasoline may easily suffer partial
combustion and misre because of its narrow ammability. This
can be veried by the quickly raised HC emissions at high excess
air ratios shown in Fig. 9a. Thus, when the k is further increased
from 1.20, BTE of the pure gasoline engine drops again with the
increase of k. As the wide ammability of hydrogen avails the
complete combustion, the hydrogen-blended gasoline engine
could be run smoothly at much higher k. This means that the
hydrogen-enriched gasoline engine could take more advantages
of lean combustion on improving BTE.

Fig. 4a. CA0-10 versus k at 1400 rpm and WOT condition.

Fig. 4b. CA10-90 versus k at 1400 rpm and WOT condition.

3.2. Combustion
It can be found from Figs. 4a and 4b that, for a given k, the
hydrogen blending results in the shortened ame development
period (CA0-10) and ame propagation period (CA10-90). This is
because hydrogen possesses lower ignition energy than gasoline
(see Table 1). Thus, the gasolineair mixtures with hydrogen

Fig. 3. Brake thermal efciency versus k at 1400 rpm and WOT condition.

blending are much easier to be ignited. Therefore, the hydrogen

blending benets shortening CA0-10 of the gasoline engine.
Besides, the high ame speed of hydrogen also contributed to the
shortened CA0-10 and CA10-90. Figs. 4a and 4b demonstrates that
the combustion durations are generally prolonged with the
increase of k due to the decreased cylinder temperature and fuel
energy ow rate.
Fig. 5 shows the central heat release position (CA50) versus k at
1400 rpm and the WOT condition. Theoretically, when the engine
adopts the MBT spark timing, CA50 should be located around 8
9 CA ATDC [43]. However, it is seen from Fig. 5 that, within the
testing range, only when k varies from 1.15 to 1.45, CA50 of the
gasoline engine blended with 3% hydrogen could be located
between 8 and 9 CA ATDC. Comparatively, CA50 of the pure gasoline engine is rstly advanced and then retarded with the increase
of k. Besides, it is also noticed from Fig. 5 that CA50 of the gasoline
engine is generally delayed from 10 CA ATDC, which is believed to
be retarded from the ideal central heat release position at the MBT
spark timing. This is because when the k is smaller than 1.15, further advancing the spark timing leads to a clear sound of knocking.
When the k is increased from 1.15, because of the decreased combustion temperature, it is allowed to advance the spark timing for
engines blended with 0% and 3% hydrogen. Thus, when the k is
greater than 1.15, CA50 of the engine blended with 3% hydrogen
could be located within the ideal position of the MBT spark timing,
and CA50 of the original gasoline engine could also be moved

S. Wang et al. / Applied Energy 136 (2014) 4350

Fig. 5. CA50 versus k at 1400 rpm and WOT condition.

Fig. 6. Qw versus k at 1400 rpm and WOT condition.

with the pure gasoline engine, the hydrogen-blended gasoline

engine could adopt more advanced spark timing at the WOT and
lean conditions.
Reducing the engine cooling loss (Qw) is important for better
BTE. Qw is calculated based on pressure data through procedures
given by Ref. [43]. The Woschni heat transfer coefcient is used
to determine the engine heat loss. Fig. 6 shows the variations of
Qw with k under lean and WOT conditions. It is seen from Fig. 6
that Qw could be reduced after hydrogen blending when k is lower
than about 1.35. But when the k is higher than 1.35, Qw is
increased with the hydrogen blending. The possible reason is that,
because of the reduced combustion durations, less fuel is burned
during the piston moving toward the bottom dead center after
the hydrogen blending. Thus, the reduced heat transfer area
enables the engine to gain dropped Qw at aH2 = 3% when the k is
lower than 1.35. However, since the relatively narrow ammability
range of gasoline leads to partial burning of the gasoline engine at
extremely high k, cylinder temperature of the original engine may
drop quickly when the k is higher than 1.35. This could be veried
by Fig. 7 which shows the cylinder temperature at exhaust valve
opening (Tevo). It is seen from Fig. 7 that Tevo of the pure gasoline
engine rst reduces with k, and when k is larger than 1.35, Tevo of
the original engine is increased with k, which is also generally
higher than that of the hydrogen-blended gasoline engine for all
tested excess air ratios. Generally, because of the reduced fuel
energy ow rate, Tevo should be decreased with the increase of
excess air ratio, and Tevo could be increased with excess air ratio
only when the post combustion occurs for the engine under lean
conditions. Thus, the raised Tevo when k is higher than 1.35 also
conrms that the original engine becomes to encounter severe post
combustion and therefore results in the quickly increased Tevo for
the original engine. Comparatively, because of the extended lean
burn limit, Tevo of the hydrogen-blended gasoline engine is
continuously decreased with the increase of excess air ratio. This
demonstrates that the combustion of hydrogengasoline blends
is faster and more stable than that of the pure gasoline at larger
excess air ratios. Therefore, when the k is larger than 1.35, Qw is
increased with hydrogen blending.
Fig. 8 depicts the coefcient of variation in indicated mean
effective pressure (CoVimep) against k. Fig. 8 demonstrates that
CoVimep of the engine blended with 3% hydrogen is lower than
that of the pure gasoline engine, especially at lean conditions. For
a given k of about 1.45, CoVimep decreases from 5.4% at aH2 = 0
to 0.9% at aH2 = 3%. This is because the engine cyclic variation is
mainly affected by the engine combustion duration [33]. As both

Fig. 7. Tevo versus k at 1400 rpm and WOT condition.

closer to the ideal position of the MBT spark timing. But when the k
is increased from 1.3, CA50 of the pure gasoline engine deviates
from the ideal position again. This is because the partial combustion and slow burning characteristics of gasoline at extremely lean
conditions make it impossible to control CA50 to be at the ideal
position under the MBT spark timing. This means that compared

47

Fig. 8. CoVimep versus k at 1400 rpm and WOT condition.

48

S. Wang et al. / Applied Energy 136 (2014) 4350

CA0-10 and CA10-90 are reduced when aH2 rises from 0% to 3%,
CoVimep of the engine blended with 3% hydrogen could be lower
than that of the original gasoline engine. Besides, it is also noticed
from Fig. 8 that CoVimep of the engine blended with 3% hydrogen
varies gently with the increase of k. Comparatively, CoVimep of the
pure gasoline engine rstly rises slowly and then elevates sharply
with the increase of k. This is because the gasoline may easily meet
the incomplete combustion and misre at extremely high k. As the
partial combustion and misre occur, the gasoline engine combustion becomes unstable which could result in the sharply raised
engine cyclic variation. Since the wide ammability range of
hydrogen avails the fully combustion of the hydrogengasoline
mixtures when k > 1.0, combustion in the engine blended 3%
hydrogen is more stable than that in the pure gasoline engine.
Therefore, CoVimep of the engine blended with 3% hydrogen varies
more gently with the k than the pure gasoline engine.

Fig. 9a gives the variations of HC emissions with k at the WOT

and lean conditions. It is seen from Fig. 9a that HC emissions are
obviously decreased with hydrogen blending for both stoichiometric and lean conditions. It can be also found from Fig. 9a that the
addition of hydrogen successfully eased the increase of HC emissions of the original gasoline engine when the excess air ratio is
larger than about 1.30. This result conrms that the hydrogen
blending has more signicant inuence on enhancing the engine
combustion and decreasing HC emissions at lean conditions than
the stoichiometric condition. This is because the hydrogen blending avails reducing the crevice effect and therefore decreasing HC
emissions. Besides, wide ammability limits of hydrogen also avail
the fully combustion of hydrogen-enriched gasolineair mixtures
at high k. This means that the hydrogen blending is able to reduce
partial burning and misre at the extremely lean conditions. Thus,
HC could be reduced after the hydrogen blending. Fig. 9a also illustrates that HC are decreased when the k is slightly raised. This is
because the properly increased k provides adequate oxygen for
the fuel to be fully burnt. But at high k, HC from the pure gasoline
engine sharply elevate with the k. Comparatively, HC of the engine
blended with 3% hydrogen could be kept at low levels at high k.
This is because the gasoline cannot be burnt completely when
the k approaches to the fuel lean burn limit. Thus, at high k, HC
from the original gasoline engine rises quickly with k due to partial
burning and incomplete combustion. Since the hydrogen blending
avails extending the engine lean burn limit, HC emissions from the

gasoline engine blended with 3% hydrogen in the intake could be

still kept at low levels when k approaches to 1.45.
Fig. 9b displays CO emission versus k at the WOT and lean conditions. It is found from Figs. 9a9c that, when the k rises from 1.00
to 1.10, CO emission is reduced sharply. This is because CO is most
sensitive to the O2 fraction in the mixtures. When the k rises from
1.00 to 1.10, the raised O2 fraction promotes the oxidation of CO
into CO2. Thus, the adoption of lean combustion is effective on
reducing CO emission. Fig. 9b also indicates that, for a specic k,
CO of the engine blended with 3% hydrogen is generally lower than
that from the pure gasoline engine. The reason is that, for a given k,
the gasoline content is reduced after hydrogen blending. This
results in the dropped carbon atom in the combustible charge,
which helps reduce CO emission. Besides, the wide ammability
limits of hydrogen also avail enhancing the combustion of hydrogengasoline blends at lean conditions. Thus, CO emission could
be reduced by hydrogen addition.
Fig. 9c depicts NOx emissions against k at the lean and WOT
conditions. Fig. 9c demonstrates that NOx are elevated with the
hydrogen blending because hydrogen has the high ame temperature and speed. Meanwhile, it is also seen from Fig. 9c that NOx
emissions are rstly raised and then reduced with the increase of
k. This is because the formation of NOx is directly related with
the O2 fraction and in-cylinder temperature. At the slightly lean
conditions, the O2 fraction is raised while the combustion temperature is not dropped obviously. Thereby, NOx are generally
enhanced at the slightly lean conditions. Comparatively, when
the k is further elevated, although the O2 fraction is raised, the cylinder temperature is decreased quickly. Thus, the further increased
excess air ratio avails decreasing NOx emissions.
Figs. 10a and 10b plot the variations of particulate number and
mass at the WOT and lean conditions, respectively. It is seen from
Figs. 10a and 10b that, for a specied k, both particulate number
and mass are reduced obviously after the hydrogen addition. At a
given k of 1.00, the particulate number and mass are decreased
by 64.2% and 53.3% after the hydrogen addition, respectively. This
is because the decreased gasoline supplement contributes to the
reduced carbon atom in the fuel. Besides, some investigations also
show that the hydrogen blending may block the polycyclic aromatic hydrocarbons growth, which is another reason for reduced
particulate emissions [44]. Furthermore, reducing the exhaust
temperature also avails reducing particulate emissions [45]. As
the combustion duration is shortened, hydrogen blending prevents
the post combustion and therefore decreases the exhaust temperature. Thus, the hydrogen blending reduces particulate emissions
from the gasoline engines.

Fig. 9a. HC emissions versus k at 1400 rpm and WOT condition.

Fig. 9b. CO emission versus k at 1400 rpm and WOT condition.

3.3. Emissions

S. Wang et al. / Applied Energy 136 (2014) 4350

49

enhanced combustion. The addition of hydrogen avails improving

the engine thermal efciency for all tested k.
The shortened CA0-10 and CA10-90 are obtained after hydrogen
blending thanks to the high burning velocity. Since the shortened
combustion duration means that less fuel is burned during the
piston downward movement, the engine cooling loss is dropped
after the hydrogen blending. CoVimep of the 3% hydrogen-blended
gasoline engine is obviously lower than that of the original engine,
especially at lean conditions. This symbolizes that the hydrogenblended gasoline engine may run at the WOT condition and control
the engine load by adjusting k.
Because of the enhanced combustion, the hydrogen blending
contributed to the reduced HC, CO and particulate emissions from
the gasoline engine. However, NOx emissions are increased after
the hydrogen addition due to the raised cylinder temperature.

Fig. 9c. NOx emissions versus k at 1400 rpm and WOT condition.

Acknowledgments
This work was supported by National Program on Key Basic
Research Project (973 Program) (Grant No. 2013CB228403), Key
Program of Sci & Tech Project of Beijing Municipal Commission of
Education (Grant No. KZ201210005002), Ph.D. Programs
Foundation of Ministry of Education of China (Grant No.
20111103110010) and Beijing Municipal Natural Science Foundation (Grant No. 3122006). Authors also thank Beijing University of
Technology for the nancial support to this study.
References

Fig. 10a. Particulate number versus k at 1400 rpm and WOT condition.

Fig. 10b. Particulate mass versus k at 1400 rpm and WOT condition.

4. Conclusions
The performance of a hydrogen-blended gasoline engine at the
WOT and lean conditions was explored in this paper. The test
results conrm that the 3% hydrogen-blended gasoline produces
lower Bmep than the original gasoline engine when k is smaller
than 1.20. But at high k, Bmep of the hydrogen-blended gasoline
engine is higher than the original gasoline engine due to the

[1] Naja G, Ghobadian B, Tavakoli T, Buttsworth DR, Yusaf TF, Faizollahnejad M.

Performance and exhaust emissions of a gasoline engine with ethanol blended
gasoline fuels using articial neural network. Appl Energy 2009;86:6309.
[2] Yang HH, Liu TC, Chang CF, Lee E. Effects of ethanol-blended gasoline on
emissions of regulated air pollutants and carbonyls from motorcycles. Appl
Energy 2012;89:2816.
[3] Clairotte M, Adam TW, Zardini AA, Manfredi U, Martini G, Krasenbrink A, et al.
Effects of low temperature on the cold start gaseous emissions from light duty
vehicles fuelled by ethanol-blended gasoline. Appl Energy 2013;102:4454.
[4] Vancoillie J, Demuynck J, Sileghem L, Ginste MVD, Verhelst S, Brabant L, et al.
The potential of methanol as a fuel for ex-fuel and dedicated spark-ignition
engines. Appl Energy 2013;102:1409.
[5] Turner JEG, Pearson RJ, Dekker E, Iosefa B, Johansson K, Bergstrm K. Extending
the role of alcohols as transport fuels using iso-stoichiometric ternary blends
of gasoline, ethanol and methanol. Appl Energy 2013;102:7286.
[6] Liang C, Ji C, Gao B. Load characteristics of a spark-ignited ethanol engine with
DME enrichment. Appl Energy 2013;112:5006.
[7] He Y, Yan F, Yu H, Yuan S, Tong Z, Sheng G. Hydrogen production in a lightdriven photoelectrochemical cell. Appl Energy 2014;113:1648.
[8] Martnez I, Romano MC, Fernndez JR, Chiesa P, Murillo R, Abanades JC. Process
design of a hydrogen production plant from natural gas with CO2 capture
based on a novel Ca/Cu chemical loop. Appl Energy 2014;114:192208.
[9] Maciel CG, Silva TF, Assaf EM, Assaf JM. Hydrogen production and purication
from the watergas shift reaction on CuO/CeO2TiO2 catalysts. Appl Energy
2013;112:529.
[10] Rossi F, Nicolini A. An experimental investigation to improve the hydrogen
production by water photoelectrolysis when cyanin-chloride is used as
sensibilizer. Appl Energy 2012;97:76370.
[11] Bartolozzi I, Rizzi F, Frey M. Comparison between hydrogen and electric
vehicles by life cycle assessment: a case study in Tuscany, Italy. Appl Energy
2013;103:10113.
[12] Tang Y, Yuan W, Pan M, Wan Z. Experimental investigation on the dynamic
performance of a hybrid PEM fuel cell/battery system for lightweight electric
vehicle application. Appl Energy 2011;88:6876.
[13] Sen AK, Wang J, Huang Z. Investigating the effect of hydrogen addition on
cyclic variability in a natural gas spark ignition engine: Wavelet
multiresolution analysis. Appl Energy 2011;88:48606.
[14] Saravanan N, Nagarajan G. Performance and emission studies on port injection
of hydrogen with varied ow rates with diesel as an ignition source. Appl
Energy 2010;87:221829.
[15] Shin B, Cho K, Han D, Song S, Chun KM. Hydrogen effects on NOx emissions and
brake thermal efciency in a diesel engine under low-temperature and heavyEGR conditions. Int J Hydrogen Energy 2011;36:628191.
[16] Zhen X, Wang Y, Xu S, Zhu Y, Tao C, Xu T, et al. The engine knock analysis an
overview. Appl Energy 2012;92:62836.
[17] Li HL, Karim GA. Knock in spark ignition hydrogen engines. Int J Hydrogen
Energy 2004;29:85965.

50

S. Wang et al. / Applied Energy 136 (2014) 4350

[18] Ji C, Wang S, Zhang B. Performance of a hybrid hydrogengasoline engine

under various operating conditions. Appl Energy 2012;97:5849.
[19] Liu X, Ji C, Gao B, Wang S, Liang C, Yang J. A laminar ame speed correlation of
hydrogenmethanol blends valid at engine-like conditions. Int J Hydrogen
Energy 2013;38:155009.
[20] Christodoulou F, Megaritis A. Experimental investigation of the effects of
separate hydrogen and nitrogen addition on the emissions and combustion of
a diesel engine. Int J Hydrogen Energy 2013;38:1012640.
[21] Wang J, Huang Z, Tang C, Zheng. Effect of hydrogen addition on early ame
growth of lean burn natural gasair mixtures. Int J Hydrogen Energy
2010;35:724652.
[22] Wang J, Huang Z, Zheng J, Miao H. Effect of partially premixed and hydrogen
addition on natural gas direct-injection lean combustion. Int J Hydrogen
Energy 2009;34:923947.
[23] He Y, Ma F, Deng J, Shao Y, Jian X. Reducing the idle speed of an SI CNG engine
fueled by HCNG with high hydrogen ratio. Int J Hydrogen Energy
2012;37:8698703.
[24] Lata DB, Misra A. Analysis of ignition delay period of a dual fuel diesel engine
with hydrogen and LPG as secondary fuels. Int J Hydrogen Energy
2011;36:374656.
[25] Lata DB, Misra A, Medhekar S. Investigations on the combustion parameters of
a dual fuel diesel engine with hydrogen and LPG as secondary fuels. Int J
Hydrogen Energy 2011;36:1380819.
[26] Lata DB, Misra A, Medhekar S. Effect of hydrogen and LPG addition on the
efciency and emissions of a dual fuel diesel engine. Int J Hydrogen Energy
2012;37:608496.
[27] Lata DB, Misra A. Theoretical and experimental investigations on the
performance of dual fuel diesel engine with hydrogen and LPG as secondary
fuels. Int J Hydrogen Energy 2010;35:1191831.
[28] Diguez PM, Urroz JC, Sdaba SM, Ezcurdia AP, Amurrio MB, Sinz D, et al.
Experimental study of the performance and emission characteristics of an
adapted commercial four-cylinder spark ignition engine running on
hydrogenmethane mixtures. Appl Energy 2014;113:106876.
[29] Ji C, Wang S, Zhang B. Effect of spark timing on the performance of a hybrid
hydrogengasoline engine at lean conditions. Int J Hydrogen Energy
2010;35:34655.
[30] Wang S, Ji C, Zhang B. Combustion and emissions characteristics of a hybrid
hydrogengasoline engine under various loads and lean conditions. Int J
Hydrogen Energy 2010;35:571422.

[31] Navarro E, Leo TJ, Corral R. CO2 emissions from a spark ignition engine
operating on natural gashydrogen blends (HCNG). Appl Energy
2013;101:11220.
[32] Ceviz MA, Sen AK, Kleri AK, ner IV. Engine performance, exhaust emissions,
and cyclic variations in a lean-burn SI engine fueled by gasolinehydrogen
blends. Appl Therm Eng 2012;36:31424.
[33] Wang S, Ji C. Cyclic variation in a hydrogen-enriched spark-ignition gasoline
engine under various operating conditions. Int J Hydrogen Energy
2012;37:11129.
[34] Ji C, Wang S, Zhang B, Liu X. Emissions performance of a hybrid hydrogen
gasoline engine-powered passenger car under the new European driving cycle.
Fuel 2013;106:8735.
[35] Fontana G, Galloni E. Experimental analysis of a spark-ignition engine using
exhaust gas recycle at WOT operation. Appl Energy 2010;87:218793.
[36] Wei H, Zhu T, Shu G, Tan L, Wang Y. Gasoline engine exhaust gas recirculation
a review. Appl Energy 2012;99:53444.
[37] Xie F, Li X, Wang X, Su Y, Hong W. Research on using EGR and ignition timing
to control load of a spark-ignition engine fueled with methanol. Appl Therm
Eng 2013;50:108491.
[38] Szwaja S, Bhandary KR, Naber JD. Comparisons of hydrogen and gasoline
combustion knock in a spark ignition engine. Int J Hydrogen Energy
2007;32:507687.
[39] Szwaja S, Jamrozik A, Tutak W. A two-stage combustion system for burning
lean gasoline mixtures in a stationary spark ignited engine. Appl Energy
2013;105:27181.
[40] Verhelst S, Wallner T. Hydrogen-fueled internal combustion engines. Prog
Energy Combust Sci 2009;35:490527.
[41] Kondo S, Takizawa K, Takahashi A, Tokuhashi K, Sekiya A. A study on
ammability limits of fuel mixtures. J Hazard Mater 2008;155:4008.
[42] Le Chatelier H, Boudouard O. Limits of ammability of gaseous mixtures. Bull
Soc Chim (Paris) 1898;19:4838.
[43] Heywood JB. Internal combustion engine fundamentals. New York: McGrawHill; 1988.
[44] Guo H, Liu F, Smallwood GJ, Gulder OL. Numerical study on the inuence of
hydrogen addition on soot formation in a laminar ethyleneair diffusion ame.
Combust Flame 2006;145:32438.
[45] Ning Z, Cheung CS, Liu SX. Experimental investigation of the effect of exhaust
gas cooling on diesel particulate. J Aerosol Sci 2004;35:33345.