Energy Conversion and Management 114 (2016) 68–74

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Performance of a hydrogen-enriched ethanol engine at unthrottled

and lean conditions
Bo Zhang, Changwei Ji ⇑, Shuofeng Wang
College of Environmental and Energy Engineering, Key Laboratory of Beijing on Regional Air Pollution Control and Collaborative Innovation Center of Electric Vehicles in Beijing,
Beijing University of Technology, Beijing 100124, China

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

Article history: Concerning the throttling loss under part load conditions, it is feasible to further improve the engine ther-
Received 17 November 2015 mal efficiency through operating the engine under the unthrottled condition and controlling its load by
Accepted 31 January 2016 changing the excess air ratio. However, the narrow flammability of ethanol may lead the ethanol engine
Available online 16 February 2016
to encounter high cyclic variations under unthrottled and lean conditions. The addition of hydrogen is
potentially helpful for solving this problem. In this test, the engine was run under an speed of
Keywords: 1400 rpm and unthrottled conditions. The hydrogen volume fractions in the intake were respectively
Hydrogen
kept at 0% and 3%. For a given hydrogen blending level, the ethanol flow rate was reduced to enable
Ethanol
Combustion
the engine to run under lean conditions. The results showed that the engine efficiency was improved with
Emissions the blending of hydrogen. The highest thermal efficiency was improved by 6.07% after blending 3% hydro-
Unthrottled gen to the intake air. The addition of hydrogen could increase the engine torque output at lean conditions.
Both cooling and exhaust losses were decreased after the hydrogen enrichment while adopting the lean
combustion strategy. The hydrogen addition contributed to the extended lean burn limit and decreased
cyclic variation under lean conditions. HC and CO emissions were decreased whereas NOx emissions
were increased after the blending of hydrogen.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction the application of ethanol to spark-ignition engines is still limited

by its high latent heat and low energy density. Generally, the high
Concerning pressures on environmental protection and nation latent heat of ethanol avails increasing the engine volumetric effi-
energy safety, searching green fuel candidates for internal combus- ciency due to its cooling effect. However, the high latent heat of
tion engines has gained more attentions nowadays. Ethanol, which ethanol brings an obstacle for the evaporation and mixing of etha-
could be made from biomasses, is regarded as a feasible fuel that nol with air. Thus, the ethanol engine would meet difficulties in the
could be combusted in spark-ignition engines [1,2]. Compared cold start and prolonged combustion duration under low loads
with gasoline, the ethanol engine could apply larger compression [10].
ratios because of its good anti-knock characteristic [3,4]. This Ambrós et al. [9] confirmed that the fuel burning velocity could
enables the ethanol engine to gain higher thermal efficiency. be elevated by adding fuels which have faster burning characteris-
Besides, the oxygen atom in ethanol avails the clean and complete tic. Compared with ethanol, the hydrogen has smaller ignition
combustion of ethanol–air mixtures, which contributes to the energy and higher diffusive and burning speeds. Besides, the wide
reduced CO and HC emissions from ethanol engines. Moreover, flammability of hydrogen is also helpful for extending the engine
the hydroxyl bond in ethanol makes the ethanol gain higher lami- lean limit. Therefore, considering the good properties of hydrogen,
nar burning velocity than gasoline [5]. Therefore, many researches blending the hydrogen to engine could be beneficial for improving
have proved that the ethanol and ethanol–gasoline fueled engines the combustion of hydrocarbon fuels. Zhang et al. [11] found that
are capable of gaining better fuel economy and smaller harmful the ethanol laminar flame speed was increased with the hydrogen
emissions than pure gasoline engines [6–9]. addition, particularly at equivalence ratios smaller than 1.4.
Although there are many good physicochemical properties of Besides, the addition of hydrogen also contributed to the improved
ethanol which help improve the combustion of ethanol engines, flame stability due to the shortened Markstein length.
Al-Hamamre et al. [12] confirmed that the fuel laminar flame
⇑ Corresponding author. Tel./fax: +86 1067392126. speed was rose linearly when the hydrogen blending ratio was
E-mail address: [email protected] (C. Ji). smaller than 60%. Investigations done by Huang et al. [13–16]

http://dx.doi.org/10.1016/j.enconman.2016.01.073
0196-8904/Ó 2016 Elsevier Ltd. All rights reserved.
 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74 69

Nomenclature

Bmep brake mean effective pressure (bar) Tevo average cylinder temperature at the exhaust valve
CA crank angle opening (K)
CA0–10 crank angle duration from spark discharge to 10% heat HC hydrocarbon
release of the total fuel (°CA) CO carbon monoxide
CA10–90 crank angle duration from 10% to 90% heat release of the NOx nitrogen oxides
total fuel (°CA) k global excess air ratio of the hydrogen–gasoline–air
CoVimep coefficient of variation in indicated mean effective pres- mixture
sure (%) aH2 hydrogen volume fraction in the intake (%)
HECU hybrid electronic control unit
Qw cooling loss (J/cycle)

demonstrated that the addition of hydrogen helped increase H, O the combustion process, reducing the pumping loss of spark-
and OH radicals during the combustion, which were effective on ignition engine at the part load condition is another challenge for
enhancing the flame stability and burning velocity of fuel–air mix- improving the engine fuel economy. In recent years, decreasing
tures. They also concluded that, by elevating the laminar flame the engine pumping loss at part load conditions has gained many
speed, the engine cyclic variation could be eased since the flame attentions. One feasible way for decreasing pumping work is to
was less affected by the turbulent flow. These results verify the fact keep the engine to run under the unthrottled condition. At the
that the basic combustion characteristics of ethanol and hydrocar- unthrottled condition, the engine load could be control by adjust-
bon fuels could be enhanced through the hydrogen enrichment. ing the valve lift and timing [31,32]. However, this method may
Therefore, both pure hydrogen and hydrogen-blended engines result in the increased throttling loss at the intake valve. Another
could gain better performance than the conventional fuel engines. way for realizing the unthrottled operation of spark-ignition engi-
Yang et al. [17–19] conducted detailed experimental and numeri- nes is to control the exhaust gas recirculation ratio and mixture’s
cal investigations on the pure hydrogen engines. He concluded excess air ratio. Schafer [33] tried to run a hydrogen-blended
from test results that the hydrogen engine could gain higher ther- methanol engine at the unthrottled condition. The test results con-
mal efficiency than the conventional engine. By properly design the firmed the fact that the engine gained better efficiency and shorter
operating strategy, the abnormal combustion could be eased. combustion duration by the hydrogen addition. Vancoillie et al.
Saravanan et al. [20,21] investigated the effect of hydrogen addition [31] compared the hydrogen-enriched methanol engine perfor-
on the performance and combustion of diesel engines. The results mance under the unthrottled condition with two load control
showed that the engine peak pressure and thermal efficiency were strategies of lean combustion and exhaust recirculation. The
improved by hydrogen addition. Smoke emissions were reduced results demonstrated that both lean combustion and exhaust gas
after the hydrogen blending, but the addition of hydrogen resulted recirculation (EGR) could realize the load control for the
in raised NOx emissions due to the enhanced combustion temper- hydrogen-blended methanol engine under the unthrottled condi-
ature. To solve this problem, he further investigated the adoption tion. The EGR strategy gained more advantage at low load because
of exhaust gas recirculation on controlling NOx from the of the heating effect. However, the adoption of EGR could result in
hydrogen-enriched diesel engines. The results indicated that NOx the decreased flame propagation speed, which may block the fully
from hydrogen-enriched diesel engines could be controlled by combustion of fuel–air mixtures.
properly adopting exhaust gas recirculation. Park et al. [22,23] Therefore, running an engine at the unthrottled condition and
found that the hydrogen blending was effective on improving effi- controlling its load by adjusting the excess air ratio is more suitable
ciency for the engines fueled by natural gas. The hydrogen blend- for reducing the pumping loss and improving the engine efficiency.
ing was able to reduce carbon-related emissions due to the However, compared with the throttle controlled part load condi-
increased H/C ratio. To control NOx emissions caused by the raised tions, the strong turbulent flow at unthrottled condition may easily
cylinder temperature, they proposed a late ignition strategy. Fur- cause the increased cyclic variation and even partial combustion or
thermore, they also explored the influence of hydrogen- misfire for the lean fuel–air mixtures, particularly for fuels which
contained reformed gas addition on the combustion of ethanol have the narrow flammability and low burning velocity. Moreover,
engines. Their results demonstrated that hydrogen blending stim- the high latent heat of ethanol also puts barriers for running the
ulated the engine efficiency and reduced the engine cyclic varia- engine under the unthrottled and lean conditions. As the hydrogen
tion. Yousufuddin et al. [24] investigated the performance of has a wide flammability, low ignition energy and high flame prop-
hydrogen-enriched ethanol engine under different compression agation speed, the addition of hydrogen seems to be capable of
ratios and spark timings. He concluded from the experiment that improving the combustion stability of ethanol engine under
the engine brake specific fuel consumption could be reduced by unthrottled and lean conditions. However, at present, there are lim-
increasing the compression ratio. Ji et al. [25–27] developed a coor- ited published papers which have investigated the effect of hydro-
dinative control strategy for the hydrogen-enriched gasoline gen addition on the combustion and emissions characteristics of
engines. Results from the vehicle test done in chassis dynamome- ethanol engines under unthrottled and lean conditions. Therefore,
ter indicated that HC and CO were respectively decreased by over in this paper, the performance of a hydrogen-enriched ethanol
60% under the regulated testing cycles of passenger cars. Akansu engine is investigated under lean and unthrottled conditions. The
et al. [28–30] ran a natural gas engine blended with hydrogen result of this article is potentially referenced for developing new
under the wide open throttle condition. His investigation control strategies of the hydrogen-enriched ethanol engines.
confirmed the fact that the hydrogen blending was capable of
reducing CO and HC exhausted by natural gas engines under the 2. Experimental setup and procedure
wide open throttle and lean conditions.
All in all, the ability of hydrogen addition on improving the The experiment was conducted on a 1.6 L spark-ignition engine.
engine performance has been widely proved. Except for improving Before the experiment, to enable the hydrogen and ethanol to be
70 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74

injected and mixed together at the runners of each cylinder, four 3. Results and discussion
hydrogen injectors connected with a hydrogen rail were mounted
on the engine. During the experiment, the hydrogen injection pres- 3.1. Brake mean effective pressure and brake thermal efficiency
sure was kept at 3 bar. The hydrogen used in this test has a purity
of 99.999%. A hybrid electronic control unit was developed to con- Fig. 2 shows the variations of Bmep with k for engine running at
trol injection pulses of hydrogen and ethanol, as well as the spark the unthrottled condition. It is seen from Fig. 2 that, since the mix-
timing. tures’ energy is decreased with the increase of excess air ratio, rais-
Fig. 1 gives the schematic diagram of experimental system. The ing k causes the dropped Bmep for both 0% and 3% hydrogen-
engine speed is controlled by an eddy current dynamometer which enriched ethanol engines. The decreased Bmep indicates that the
was made by Power Link, the maximum measurement errors in the engine torque could be controlled in a range through adjusting
engine speed and torque are ±1 rpm and ±0.28 N m, respectively. the excess air ratio. This provides a good potential of running the
Two thermal mass flow meters are applied to measure the hydro- engine under the unthrottled condition and controlling its load
gen and air flow rates. The maximum deviations are ±0.02 L/min by changing k. Then, by applying this strategy, it is reasonable that
for hydrogen and ±0.1 L/min for air. The ethanol flow rate is mea- the engine could spend less work on pumping than controlling the
sured with a maximum deviation of ±0.33 g/min. A Kistler optical load through adjusting the opening of throttle.
encoder typed 2613B, a piezoelectric pressure sensor typed Kistler Fig. 2 also demonstrates that Bmep is reduced after the hydro-
6117 and a combustion analyzer typed Dewetron 800 are used to gen enrichment when k < 1.16. Comparatively, at the conditions of
accomplish the heat release analysis. During the experiment, the k > 1.16, Bmep of the ethanol engine is increased by hydrogen
resolution of crank angle measurement is set to be 0.2 °CA. The blending. The possible reason is that the hydrogen volume energy
emissions analyzer applied in this study is Horiba MEXA7100 density is very low compared with that of the ethanol. Besides, the
DEGR. This emissions analyzer measures HC, CO and NOx through high stoichiometric air-to-fuel ratio of hydrogen also means that
flame ionization detection, nondispersive infrared and chemilumi- the combustion of hydrogen could quickly consume more air than
nescent method, respectively. The sensitivity of HC, CO and NOx ethanol. Therefore, at a specific intake pressure, engine speed and
measurements are 1 ppm. The excess air ratio of the hydrogen–et k, the energy density of hydrogen–ethanol–air mixtures is smaller
hanol–air mixtures is monitored by a Horiba MEXA730 AF than that of the ethanol–air mixtures. Thus, at low excess air ratios,
analyzer. in view of the dropped fuel energy flow rate, the hydrogen blend-
In the experiment, the main throttle was kept at the wide open ing results in the slightly decreased engine Bmep. When the k is
to simulate the engine running at the unthrottled condition. The greater than 1.16, the narrow flammability of ethanol makes the
engine speed was controlled at 1400 rpm. The hydrogen volume combustion of pure ethanol engine become unstable at lean condi-
fractions in the total intake (aH2) gas were respectively kept at tions. Comparatively, since the addition of hydrogen is capable of
0% and 3%. For a given aH2, the mixture’s excess air ratio (k) was improving the engine combustion stability [34], the 3%
adjusted through changing the ethanol injection duration. The lar- hydrogen-blended ethanol–air mixtures could be burnt well under
gest excess air ratios for the 0% and 3% hydrogen-enriched ethanol lean conditions. Thus, Bmep of the ethanol engine reduces more
engine were 1.30 and 1.49, respectively. This is because further quickly than that of the 3% hydrogen-blended ethanol engine.
increasing the excess air ratio by reducing the ethanol injection Fig. 3 gives the variations of engine brake thermal efficiency
would cause the sharp increase in cyclic variation, and the engine (BTE) with k under unthrottled conditions. It is seen from Fig. 3
would become very unstable. For all testing points, spark timings that, the properly increased k helps improve BTE. However, BTE
for the maximum brake torque were applied. could be decreased quickly if the k is too high. This is because
the increased oxygen fraction at lean conditions avails the com-
plete combustion of fuel–air mixtures. Consequently, at properly
lean conditions, BTE of the hydrogen-blended ethanol engine could
be improved. It is also noticed from Fig. 3 that the peak BTE of pure
ethanol engine is found to be 34.6% at a k of 1.12. Comparatively,
the engine highest BTE achieves 36.7% after blending 3% hydrogen
in the intake, which appears at a k of 1.25. Moreover, when excess

Fig. 1. The schematics of the experimental system. 1. Hydrogen cylinder container.

2. Hydrogen pressure adjusting valve. 3. Hydrogen pressure meter. 4. Hydrogen
mass flow meter. 5. Backfire arrestor. 6. Hydrogen injector. 7. Throttle. 8. Air mass
flow meter. 9. Idle valve. 10. OECU. 11. HECU. 12. Calibration computer. 13. Fuel
tank. 14. Fuel mass flow meter. 15. Fuel pump. 16. Fuel injector. 17. Ignition
module. 18. Pressure transducer with a spark plug. 19. Optical encoder. 20. Charge
amplifier. 21. A/D converter. 22. Combustion analyzer. 23. O2 sensor. 24. A/F
analyzer. 25. Emissions sampling pipe. 26. Horiba MEXA-7100DEGR emissions
analyzer. (a) Signals from the OECU to the HECU. (b1) Calibration and control
signals from the calibration computer to the HECU. (b2) Data signals from the HECU
to the calibration computer. Fig. 2. Bmep versus k under the unthrottled condition.
 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74 71

Fig. 4b. CA10–90 versus k under the unthrottled condition.

Fig. 3. Brake thermal efficiency versus k under the unthrottled condition.

energy flow rate is reduced for a specific engine speed and mani-
air ratio exceeds relevant value for the best thermal efficiency, BTE
folds pressure. Since the reduced fuel energy flow rate could easily
of the pure ethanol engine quickly reduces with the increase of k.
lead to the decreased cylinder temperature, the combustion dura-
However, BTE of the 3% hydrogen-enriched ethanol engine shows
tion is reasonably prolonged with the increase of k. Figs. 4a and 4b
only a little reduction when k is changed. The possible reason is
also indicate that the 3% hydrogen-enriched ethanol engine could
that the wide lean flammability limit of hydrogen enables the hy
gain shorter combustion periods than the pure ethanol engine. This
drogen–ethanol–air mixtures to accomplish a more complete com-
is because the good ignitability of hydrogen could help the fast
bustion under lean conditions than pure ethanol. This demon-
ignition of ethanol–air mixtures, and therefore shortens CA0–10
strates that the pure ethanol engine is hard to run at high excess
of the ethanol engine. Besides, the high laminar burning velocity
air ratios with good thermal efficiency. Moreover, the ignitability
of hydrogen also enables the flame of hydrogen–ethanol–air mix-
and flame propagation speed could be improved by the hydrogen
tures to propagate quickly. Thus, the blending of hydrogen could
addition due to its low ignition energy and high laminar burning
shorten CA10–90 of the ethanol engine. Moreover, Figs. 4a and
velocity. These avail decreasing the exhaust loss and help the com-
4b also confirm that hydrogen blending is more effective on reduc-
bustion of hydrogen-blended ethanol engine move closer to the
ing the engine combustion duration at high excess air ratios. This is
top dead center. Therefore, for a given k, it is reasonable that the
because the poor lean burn ability of ethanol is easy to result in
ethanol engine blended with hydrogen could gain higher BTE.
slow burning for the pure ethanol engine under lean conditions.
Comparatively, as the hydrogen could be burnt quickly and
3.2. Combustion analysis completely at very lean conditions, the 3% hydrogen-blended
ethanol–air mixtures could be combusted much quicker than the
Figs. 4a and 4b give the variations of flame development (CA0– pure ethanol–air mixtures under high ks.
10) and propagation (CA10–90) periods with k under unthrottled Fig. 5 shows the cylinder temperature at the exhaust valve
conditions. It is found from Figs. 4a and 4b that the raised k gener- opening (Tevo) against k. In this paper, Tevo is derived from the
ally results in prolonged combustion duration for hydrogen- state equation (PV = mRT) with an ideal gas assumption [35]. As
enriched ethanol engines. The proper explanation for this fact Tevo is related with the enthalpy of exhaust gas, it could be seen
could be ascribed to that the increased k means that the fuel as an indicator for the engine exhaust losses. Fig. 5 verifies that

Fig. 4a. CA0–10 versus k under the unthrottled condition. Fig. 5. Tevo versus k under the unthrottled condition.
72 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74

Tevo is reduced nearly linearly when k increases. The explanation

is that the reduced fuel energy flow rate at lean conditions leads
to the dropped combustion temperature. Moreover, it is seen from
Fig. 5 that, for a given k, Tevo is reduced after hydrogen addition.
This is explained by that the engine combustion duration shown
in Figs. 4a and 4b is shortened by the hydrogen blending. The
shortened CA0–10 and CA10–90 demonstrate that the combustion
of hydrogen–ethanol–air mixtures occurs much closer to the top
dead center. This means that the post combustion is reduced after
the addition of hydrogen. Since the combustion is getting concen-
trated around the top dead center after the addition of hydrogen, it
is reasonable to deduce that the hydrogen blending could help
reduce Tevo and exhaust loss of the ethanol engine.
Cooling loss directly affects the engine thermal efficiency. Fig. 6
plots the variations of cooling loss (Qw) with k. The Woschni equa-
tion is applied when calculates Qw. As it could be seen from Fig. 6,
because of the dropped fuel energy flow rate, Qw is generally
reduced with the increase of k. It is also found from Fig. 6 that
Fig. 7. CoVimep versus k under the unthrottled condition.
Qw is reduced by hydrogen blending. The explanation is that the
engine cooling loss is generally influenced by the combustion tem-
perature, the duration of high-temperature combustion and the
heat transfer area. Although the quick combustion and high com- ensures the stable combustion under lean conditions. Furthermore,
bustion temperature characteristics of hydrogen may result in the elevated flame propagation velocity and shortened combustion
the raised combustion temperature, the reduced combustion dura- duration after the addition of hydrogen could also be ascribed to be
tion after the hydrogen blending contributed to the shortened the reason for the dropped CoVimep. This is because the decreased
high-temperature combustion duration and dropped heat transfer combustion duration could help the fuel–air mixtures take less
areas. Therefore, Qw could be reduced by the addition of hydrogen negative effect of turbulent combustion on the engine cyclic varia-
under the testing condition. tion. Therefore, CoVimep of the ethanol engine under the lean and
Cyclic variation is generally a limitation for adopting lean com- unthrottled conditions could be controlled within lower values by
bustion strategy on commercial spark-ignition engines. Fig. 7 the hydrogen blending.
shows the coefficient of variation in indicated mean effective pres-
sure (CoVimep) against k under unthrottled conditions. It is seen
from Fig. 7 that CoVimep first rises gently and then rises obviously 3.3. Emissions characteristics
with the increase of k. This is because the influence of strong tur-
bulence within the cylinder is getting pronounced for the weak It is seen from Fig. 8 that, for the mono ethanol engine, when k
flames which are generally found for lean conditions [15,36]. Thus, is gradually raised, HC first reduce and then quickly increase. This
when the fuel–air mixtures become extremely lean, CoVimep is because the properly increasing k could provide more oxygen
could be quickly increased when the k is raised. For the pure etha- which is beneficial for the complete combustion of the ethanol.
nol engine running under the unthrottled conditions, further However, when the mixtures are further leaned, the quickly
increasing the excess air ratio from 1.30 would cause a sudden reduced cylinder temperature tends to cause the prolonged burn-
misfire and a sharp increase in CoVimep. This could be verified ing duration and incomplete combustion of ethanol–air mixtures.
by the quickly increased HC emissions when the k exceeds 1.20. Therefore, when the k is larger than 1.20, HC from the ethanol
Comparatively, for the 3% hydrogen-enriched ethanol engine, the engine quickly increase with the increase of k. Fig. 8 also shows
engine could run stably even when the k reaches about 1.5. This that, for a specific k, the addition of hydrogen is effective on reduc-
is because the hydrogen addition could extend the lean limit which ing HC emissions, due to the reduced C/H ratio of combustible

Fig. 6. Qw versus k under the unthrottled condition. Fig. 8. HC versus k under the unthrottled condition.
 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74 73

Fig. 9. CO versus k under the unthrottled condition. Fig. 10. NOx versus k under the unthrottled condition.

mixtures after the hydrogen blending. Moreover, the flame 4. Conclusions

quenching near the wall and cylinder head may also results in
HC emissions. Particularly under lean conditions, the quenching This paper experimentally investigated the effect of hydrogen
effect may generally cause more HC emissions. Compared with addition on the combustion and emissions characteristics of an
ethanol, the hydrogen possesses a shorter quenching distance, ethanol engine under 1400 rpm, lean and unthrottled conditions.
combined with the enhanced combustion and extended lean burn The main conclusions are listed as follows:
limit after the hydrogen blending, HC emissions caused by the
quenching effect could be reduced effectively through hydrogen 1. The addition of hydrogen is effective on reducing the engine
enrichment. It is further noticed from Fig. 8 that HC emissions from cyclic variation and extending the lean burn limit due to the
the 3% hydrogen-blended ethanol engine vary smoothly and wide flammability and high flame speed of hydrogen. This pro-
increase gently with k. This is because the wide flammability and vides a possibility to run the engine under the unthrottled con-
high flame speed of hydrogen enable the hydrogen–ethanol–air ditions and control its load by adjusting the excess air ratio.
blends to be burnt completely under lean conditions. Therefore, 2. The 3% hydrogen-blended ethanol engine could obtain higher
HC emissions did not ‘‘jump up” at high ks. Except for combustion brake thermal efficiency than the pure ethanol engine. The
process, HC emissions are also influenced by oxidation process in addition of hydrogen helps improve brake mean effective pres-
the tailpipe. As it is shown previously, for a given hydrogen blend- sure only at lean conditions.
ing level, Tevo is gradually decreased with the increase of k. The 3. The addition of hydrogen contributes to shortened combustion
dropped Tevo at lean conditions blocks the oxidation of HC in duration, the decreased cooling and exhaust losses for the etha-
the tailpipe. Therefore, HC emissions from the 3% hydrogen- nol engine under the unthrottled condition.
enriched ethanol engine are slightly raised with k. 4. Under the lean and unthrottled conditions, HC and CO emis-
Fig. 9 gives the variations of CO emission with k for the 0% and sions are reduced, but NOx emissions are increased after the
3% hydrogen-enriched ethanol engines under 1400 rpm and hydrogen addition.
unthrottled conditions. CO from both of the pure ethanol or 3%
hydrogen-enriched ethanol engines are decreased when k is grad-
ually raised. The possible reason for this result could be ascribed to Acknowledgments
the fact that CO is mostly influenced by the oxygen level within the
cylinder. Under lean conditions, the oxygen fraction in the fuel–air This work was supported by National Natural Science Founda-
mixtures is increased. The raised oxygen fraction helps reduce the tion (Grant No. 51476002), National Key Basic Research Develop-
rich zones within the combustion chamber and therefore avails ment Project (973) (Grant No. 2013CB228403), Beijing Municipal
reducing CO emissions. Fig. 9 also demonstrates that the hydrogen Commission of Science and Technology (Grant No.
blending results in the decreased CO at a specific k. This is because Z141100003814017) and Sci & Tech Project of Beijing Municipal
the decreased C/H ratio after hydrogen addition contributes to the Commission of Education (Grant No. KM201510005011).
reduced C-related emissions for the hydrogen-enriched engines.
Fig. 10 plots the variations of NOx emissions with k under
unthrottled conditions. It is found from Fig. 10 that, at a given References
hydrogen blending level, NOx emissions are gradually reduced
[1] Baêta JGC, Pontoppidan M, Silva TRV. Exploring the limits of a down-sized
with the increase of k. This is because the formation of NOx is ethanol direct injection spark ignited engine in different configurations in
mainly influenced by the cylinder temperature and oxygen level. order to replace high-displacement gasoline engines. Energy Convers Manage
At lean conditions, although oxygen fraction is increased, the 2015;105:858–71.
[2] Munsin R, Laoonual Y, Jugjai S, Matsuki M, Kosaka H. Effect of glycerol
dropped cylinder temperature provides a positive condition for ethoxylate as n ignition improver on injection and combustion characteristics
suppressing the formation of NOx. Thus, NOx are obviously of hydrous ethanol under CI engine condition. Energy Convers Manage
reduced at lean conditions. Since the hydrogen has a higher adia- 2015;98:282–9.
[3] Huang Y, Hong G, Huang R. Numerical investigation to the dual-fuel spray
batic flame temperature than the ethanol, NOx emissions are
combustion process in an ethanol direct injection plus gasoline port injection
raised after the hydrogen addition. (EDI + GPI) engine. Energy Convers Manage 2015;92:275–86.
74 B. Zhang et al. / Energy Conversion and Management 114 (2016) 68–74

[4] Liang C, Ji C, Gao B, Liu X, Zhu Y. Investigation on the performance of a spark- [20] Saravanan N, Nagarajan G. An experimental investigation on manifold-
ignited ethanol engine with DME enrichment. Energy Convers Manage injected hydrogen as a dual fuel for diesel engine system with different
2012;58:19–25. injection duration. Int J Energy Res 2009;33:1352–66.
[5] Jia LW, Shen MQ, Wang J, Lin MQ. Influence of ethanol–gasoline blended fuel [21] Saravanan N, Nagarajan G. Experimental investigation on a DI dual fuel engine
on emission characteristics from a four-stroke motorcycle engine. J Hazard with hydrogen injection. Int J Energy Res 2009;33:295–308.
Mater 2005;23:29–34. [22] Park C, Kim C, Choi Y, Won S, Moriyoshi Y. The influences of hydrogen on the
[6] Boretti A. Towards 40% efficiency with BMEP exceeding 30 bar in directly performance and emission characteristics of a heavy duty natural gas engine.
injected, turbocharged, spark ignition ethanol engines. Energy Convers Int J Hydrogen Energy 2011;36:3739–45.
Manage 2012;57:154–66. [23] Park C, Choi Y, Kim C, Oh S, Lim G, Moriyoshi Y. Performance and exhaust
[7] Al-Hasan M. Effect of ethanol–unleaded gasoline blends on engine emission characteristics of a spark ignition engine using ethanol and ethanol-
performance and exhaust emission. Energy Convers Manage reformed gas. Fuel 2010;89:2118–25.
2003;44:1547–61. [24] Yousufuddina S, Masood M. Effect of ignition timing and compression ratio on
[8] Chen R, Nishida K. Spray evaporation of ethanol–gasoline-like blend and the performance of a hydrogen–ethanol fuelled engine. Int J Hydrogen Energy
combustion of ethanol–gasoline blend injected by hole-type nozzle for direct- 2009;34:6945–50.
injection spark ignition engines. Fuel 2014;135:263–73. [25] Ji C, Wang S, Zhang B, Liu X. Emissions performance of a hybrid hydrogen–
[9] Ambrós WM, Lanzanova TDM, Fagundez JLS, Sari RL, Pinheiro DK, Martins MES, gasoline engine-powered passenger car under the New European Driving
et al. Experimental analysis and modeling of internal combustion engine Cycle. Fuel 2013;106:873–5.
operating with wet ethanol. Fuel 2015;158:270–8. [26] Ji C, Wang S. Effect of hydrogen addition on lean burn performance of a spark-
[10] Sales LCM, Sodré JR. Cold start characteristics of an ethanol-fuelled engine ignited gasoline engine at 800 rpm and low loads. Fuel 2011;90:1301–4.
with heated intake air and fuel. Appl Therm Eng 2012;40:198–201. [27] Ji C, Wang S, Zhang B. Performance of a hybrid hydrogen–gasoline engine
[11] Zhang Z, Li G, Ouyang L, Pan Z, You F, Gao X. Experimental determination of under various operating conditions. Appl Energy 2012;97:58458–9.
laminar burning velocities and Markstein lengths for 75% hydrous-ethanol, [28] Kahraman N, Çeper B, Akansu SO, Aydin K. Investigation of combustion
hydrogen and air gaseous mixtures. Int J Hydrogen Energy characteristics and emissions in a spark-ignition engine fuelled with natural
2011;36:13194–206. gas–hydrogen blends. Int J Hydrogen Energy 2009;34:1026–34.
[12] Al-Hamamre Z, Yamin J. The effect of hydrogen addition on premixed laminar [29] Akansu SO, Kahraman N, Çeper B. Experimental study on a spark ignition
acetylene–hydrogen–air and ethanol–hydrogen–air flames. Int J Hydrogen engine fuelled by methane–hydrogen mixtures. Int J Hydrogen Energy
Energy 2013;38:7499–509. 2007;32:4279–84.
[13] Wang J, Huang Z, Tang C, Zheng J. Effect of hydrogen addition on early flame [30] Akansu SO, Dulger Z, Kahraman N, Veziroǧlu TN. Internal combustion engines
growth of lean burn natural gas–air mixtures. Int J Hydrogen Energy fueled by natural gas—hydrogen mixtures. Int J Hydrogen Energy
2010;35:7246–52. 2004;29:1527–39.
[14] Wang J, Huang Z, Tang C, Miao H, Wang X. Numerical study of the effect of [31] Vancoillie J, Sileghem L, Ginste VM. Experimental evaluation of lean-burn and
hydrogen addition on methane–air mixtures combustion. Int J Hydrogen EGR as load control strategies for methanol engines. SAE paper 2012-01-1283;
Energy 2009;34:1084–96. 2012.
[15] Huang B, Hu E, Huang Z, Zheng J, Liu B, Jiang D. Cycle-by-cycle variations in a [32] Xie Z, Zhang J, Chang Y. Study on intake performance of unthrottled SI engine
spark ignition engine fueled with natural gas–hydrogen blends combined with with fully variable hydraulic valve actuating system. SAE paper 2014-01-2872;
EGR. Int J Hydrogen Energy 2009;34:8405–14. 2014.
[16] Pan L, Hu E, Meng X, Zhang Z, Huang Z. Kinetic modeling study of hydrogen [33] Schafer F. An investigation of the addition of hydrogen to methanol on the
addition effects on ignition characteristics of dimethyl ether at engine- operation of an unthrottled Otto engine. SAE paper 810776; 1981.
relevant conditions. Int J Hydrogen Energy 2015;40:5221–35. [34] Ma F, Wang Y, Liu H, Li Y, Wang J, Ding S. Effects of hydrogen addition on cycle-
[17] Yang Z, Wang L, Li S. Investigation into the optimization control technique of by-cycle variations in a lean burn natural gas spark-ignition engine. Int J
hydrogen-fueled engines based on genetic algorithms. Int J Hydrogen Energy Hydrogen Energy 2008;33:823–31.
2008;33:6780–90. [35] Ma F, Wang Y, Liu H, Li Y, Wang J, Zhao S. Experimental study on thermal
[18] Yang Z, Si A, Wang F, Guo N. Research into the formation process of hydrogen– efficiency and emission characteristics of a lean burn hydrogen enriched
air mixture in hydrogen fueled engines based on CFD. Int J Hydrogen Energy natural gas engine. Int J Hydrogen Energy 2007;32:5067–75.
2010;35:3051–7. [36] Wang J, Huang Z, Miao H, Wang X, Jiang D. Study of cyclic variations of direct-
[19] Yang Z, Wang L, He M, Cao Y. Research on optimal control to resolve the injection combustion fueled with natural gas–hydrogen blends using a
contradictions between restricting abnormal combustion and improving constant volume vessel. Int J Hydrogen Energy 2008;33:7580–91.
power output in hydrogen fueled engines. Int J Hydrogen Energy
2012;37:774–82.