Energy 153 (2018) 248e255

Contents lists available at ScienceDirect

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

An exploration of utilizing low-pressure diesel injection for natural

gas dual-fuel low-temperature combustion
Heping Song a, Changpeng Liu a, Yanfei Li a, *, Zhi Wang a, b, Longfei Chen c, **, Xin He a,
Jianxin Wang a
a
State Key Laboratory of Automotive Safety and Energy, Tsinghua University, Beijing 100084, China
b
Collaborative Innovation Center of Intelligent New Energy Vehicle, Beijing, 100084, China
c
School of Energy and Power Engineering, Beihang University, Beijing, 100191, China

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

Article history: It has been widely reported that natural gas dual-fuel combustion (DFC) can achieve much lower soot
Received 31 August 2017 emissions in contrast to conventional diesel combustion (CDC). Thus, using low-pressure direct injection
Received in revised form (LPDI) systems could be an alternative for current high-pressure common rail injection systems, which
13 March 2018
would significantly reduce the system cost. The present study aimed at exploring the feasibility of LPDI
Accepted 7 April 2018
Available online 9 April 2018
(low to 200 bar) for natural gas DFC in combination of the advanced low temperature combustion
technology. The comparative study between natural gas DFC and CDC were carried out. For natural gas
DFC, larger advanced injection timing was used to realize low temperature combustion and achieve long
Keywords:
Low-pressure diesel injection
ignition delay in order to counteract the negative impact of relatively poor atomization quality caused by
Low temperature combustion the low injection pressure. At DFC mode, higher CO and THC emissions were observed compared to CDC
Natural gas in the cases without EGR. However, DFC was much less sensitive to EGR rate and injection pressure.
Dual-fuel Natural gas DFC could break the trade-off between NOx and soot emissions, which could achieve low
Trade-off soot and NOx emissions (lower than Europe VI standard: 0.4 g/kW$h) simultaneously at the 42% EGR rate
and the 200 bar injection pressure.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction engines and have great advantages in the improvement in mixture

quality [8] and PM emissions reduction [9]. At the same time, the
Compression ignition (CI) diesel engines can achieve relatively high cost of CRI systems is being a concern for automotive in-
higher thermal efficiencies in comparison with spark ignition gas- dustries. Table 1 summarized the costs of different fuel injection
oline engines, which are widely used in vehicles and engineering systems [10]. The cost of a set of typical CRI system is approximately
machines [1]. However, due to the more and more stringent auto- twice as high as that of a center-mount gasoline direct injection
motive emission regulations, conventional diesel engines face great (GDI) system, which accounts for a large percentage of the entire
challenges in NOx and particulate matter (PM) emissions. There- engine cost.
fore, great efforts have been made on advanced technologies for CI Some studies focusing on the application of low-pressure in-
engines, including exhaust after-treatments [2], fuel design [3], jection strategies into dual-fuel combustion (DFC) have been car-
exhaust gas recirculation (EGR) [4] and new combustion modes, ried out. Liu et al. [11] studied the effects of diesel injection pressure
such as homogeneous charge compression ignition (HCCI) [5], (700e1400 bar) on diesel-methanol DFC and reported that CO and
premixed charge compression ignition (PCCI) [6] and reactivity HC emissions decreased while CO2 and NO2 emissions slightly
controlled compression ignition (RCCI) [7]. Besides, high-pressure increased with increasing injection pressure. Walker et al. [12] used
common rail injection (CRI) systems have been widely used in CI a set of GDI fuel supply system to inject diesel to realize RCCI. The
experimental results showed that the GDI system (injection pres-
sure: 150e200 bar) could achieve comparable engine performance
* Corresponding author. and emissions by utilizing an optimized piston in comparison with
** Corresponding author. the CRI system. Chang et al. [13] conducted the experiments to
E-mail addresses: [email protected] (Y. Li), [email protected] investigate diesel-gasoline DFC with low-pressure injection. They
(L. Chen).

https://doi.org/10.1016/j.energy.2018.04.041
0360-5442/© 2018 Elsevier Ltd. All rights reserved.
 H. Song et al. / Energy 153 (2018) 248e255 249

NOx emissions and promote efficiency simultaneously, but with

Abbreviations injection pressures higher than 500 bar [17].
Based on aforementioned discussion, a novel and practicable
AC absorption coefficient way is proposed to realize low-cost but high-efficiency and clean
ATDC after top dead center combustion for natural gas by using low-cost LPDI systems (e.g.
CA crank angel within the capability of modern GDI systems) in combination with
CNG compressed natural gas dual-fuel and low-temperature combustion. On one hand, the po-
DFC dual-fuel combustion tential increase in soot emissions due to LPDI could be counteracted
GDI gasoline direct injection by the low soot emissions of natural gas combustion to some
IMEP indicated mean effective pressure extent. On the other hand, more advanced injection timing to
LPDI low-pressure diesel injection realize LTC can also cause long ignition delay, providing more time
PM particulate matter for fuel mixture preparation and ultimately mitigating the negative
PPRR peak pressure rise rate effect of poor atomization caused by LPDI. Thus, a preliminary study
AHRR apparent heat release rate was conducted to examine the effectiveness of natural gas dual-fuel
BTDC before top dead center low-temperature combustion with a LPDI strategy.
CDC conventional diesel combustion
CRI common rail injection 2. Experimental setup
EGR exhaust gas recirculation
HPDI high-pressure direct injection The experiments were conducted using a single-cylinder engine
LHV lower heating value which was retrofitted from a mass production six-cylinder engine
LTC low temperature combustion produced by Dongfeng Motor Corporation. The engine specifica-
PN particle number tions are listed in Table 2. Fig. 1 shows the schematic of the engine
test bench. The intake air mass flow was controlled and measured
by a six-channel sonic nozzle system. Two surge tanks were used to
minimize the intake and exhaust pressure fluctuation. EGR rate was
controlled by adjusting exhaust pressure and EGR valves. Com-
Table 1
pressed natural gas (CNG) was injected into the intake manifold.
Costs of different 4-cylinder fuel injection systems [10].
Pilot fuel was injected using a CRI system. Horiba MEXA-7200 was
Injection system Cost ($) used to measure the engine-out regulated emissions including NOx,
Port fuel injection 32 CO, THC, and CO2. AVL 439 opacimeter was used to measure ab-
Side-mount GDI 227 sorption coefficient (AC), which represented soot emissions.
Center-mount GDI 365 DMS500 was used to measure PM size distribution. The in-cylinder
Center-mount CRI 707
pressure was measured with a Kistler 6125C pressure transducer.
The engine was controlled by a LabVIEW-based engine control unit,
which also integrated the function of data acquisition and com-
reported that the DFC mode with relatively low injection pressures bustion analysis.
(600e1000 bar) could achieve the same level of soot emissions as The used CNG composition in this study is shown in Table 3,
conventional diesel combustion (CDC) with relatively high injec- which was measured by the gas chromatography (Agilent 7890B).
tion pressures (1200e1400 bar). Ryu et al. [14] investigated the The diesel used in the experiment meets the Beijing V diesel fuel
effects of pilot injection pressure (300e1500 bar) on biodiesel- standard. CA10, CA50, and CA90 are defined as the crank angel (CA)
natural gas DFC. The experimental results showed that smoke where 10%, 50%, and 90% of the total heat release occur, respec-
emissions increased as the pilot injection pressure decreased, while tively. Combustion duration is defined as the CA interval between
NOx emissions showed the opposite trend. The aforementioned CA10 and CA90. Substitution ratio is defined as the ratio of the CNG
studies present the great potential to use low-pressure diesel in- lower heating value (LHV) to the total fuel LHV per cycle. EGR rate is
jection (LPDI) strategy in DFC, but the main obstacle for further calculated as the rate of CO2 mole fraction between intake and
decrease in injection pressure is the mixture quality. Low injection exhaust manifolds. The indicated thermal efficiency (ITE) was
pressure could lead to the poor atomization quality and fuel-air defined according to the following formula:
mixture, which would deteriorate the combustion and emissions,
especially for soot emissions. hi ¼ Wi =ðmd  LHVd þ mn  LHVn Þ (1)
In recent years, many research institutes have studied some key
control parameters on natural gas DFC, such as EGR fraction [15], Where hi is the ITE, Wi is the indicated work, md is the diesel
compression ratio [16], pilot fuel type [17], and injection strategy consumed per cycle, LHVd is the diesel lower heating value, mn is
[18], etc. The high-pressure direct injection (HPDI) system devel- the natural gas consumed per cycle, and LHVn is the natural gas
oped by Westport Innovations Inc. could simultaneously achieve lower heating value. The uncertainties of measured values are
high efficiency and low emissions [19]. A consensus has been
achieved that natural gas DFC mode has much lower soot emissions Table 2
compared to CDC [17,20]. Specifications of the single-cylinder engine.
Besides, low temperature combustion (LTC) is usually based on
Parameter Value
port injection and early direct injection to achieve relatively ho-
mogenous fuel-air mixture before ignition. For LTC, a large fraction Displacement 1.85 L
Compression ratio 17.3
of the fuel burns in premixed combustion phase and the combus- Bore 123 mm
tion temperature is lower than conventional CI combustion [21]. Stroke 156 mm
Resultantly, much lower NOx and soot emissions can be achieved Swirl ratio 1.85
by LTC. Some studies have used advanced direct injection timing to Direct injection pressure 0-1600 bar
CNG injection pressure 5 bar
realize natural gas dual-fuel LTC, which could significantly reduce
250 H. Song et al. / Energy 153 (2018) 248e255

deteriorate combustion. The injection timings for both CDC and

DFC modes were fixed while the injection duration varied with
injection pressure. Compared with conventional DFC, more
advanced injection timing was utilized to achieve LTC.

3. Results and discussion

In this section, the results of DFC and CDC modes conducted at

different pilot injection pressures and EGR rates were shown. A
detailed comparison between these two modes was performed in
terms of combustion and emissions characteristics. Fig. 2(a) and (b)
show the curves for the in-cylinder pressure and apparent heat
release rate (AHRR) of CDC. Three peaks in heat release were
Fig. 1. Schematic of the engine test bench.
observed. Peak heat release rate decreased with decreasing injec-
tion pressure, because lower injection pressure could lead to rela-
Table 3 tively poor fuel atomization and less ignitable fuel-air mixture, i.e.
The composition of the tested CNG. less pre-mixed combustion. Fig. 2(c) and (d) show the in-cylinder
pressure and AHRR of natural gas DFC. Natural gas DFC exhibited
Component Mole fraction (%)
single peak heat release and the heat release of pilot diesel fuel and
Methane 93.78
natural gas cannot be clearly discriminated. For DFC, the timing for
Ethane 4.91
Propane 0.79 the pilot injection was fixed at 42 CA before top dead center (BTDC)
Butane 0.18 and this early injection timing resulted in a long ignition delay and
Others 0.34 resultantly more time for evaporation. It is worth noting that the
ignition delay became shorter as the injection pressure decreased.
It has close relationship with the spray and atomization process. It
shown in Table 4. An error analysis for the calculated values, such as is well known that the increase in injection pressure could increase
ITE, was carried out based on the root mean square method [22]. the fuel penetration, promote atomization [23] and enhance the
For function R (X1, X2, … Xn), the uncertainty calculated using the utilization of the in-cylinder air [24]. However, in the present cases
following formula: with much advanced injection timing (42 CA BTDC) and small fuel
amount (10% pilot injection), the longer spray penetration length,
" better fuel atomization and faster droplet evaporation could retard
2  2  2 #12
vR vR vR ignition delay, i.e. leading to over-mixing of the pilot fuel with air
UR ¼ U þ U þ…þ U (2) (lean local equivalence ratio), reduced local temperature and finally
vx1 X1 vx2 X2 vxn Xn
the longer ignition delay.
Where x1, x2 and xn represent independent variables with The results of CA50, combustion duration, and PPRR are shown
measured uncertainties UX1 , UX2 ; and UXn , respectively. The un- in Fig. 3. The combustion phasing was retarded by EGR addition. It
certainties of calculated values are also shown in Table 4. can be attributed to the factor that EGR addition would reduce in-
Table 5 lists the experimental conditions. In this study, all ex- cylinder oxygen fraction and temperature, both of which could
periments were conducted at indicated mean effective pressure suppress combustion. CA50 generally advanced when injection
(IMEP) of about 12 bar. The combined heating value of CNG and pressure decreased for dual-fuel mode due to the shorter ignition
diesel was maintained constant. The actual IMEP slightly changed delay at lower injection pressures, as shown in Fig. 2(c) and (d).
due to the differences in efficiency. Both CDC and natural gas DFC However, the change in CA50 was fairly slight for CDC within the
experiments were performed with the EGR rates of 0% and 42%. For given range of injection pressures. Combustion duration generally
0% EGR rate, the intake pressure was 1.58 bar and equivalence ratio decreased as the injection pressure increased for both. The com-
was 0.51. It was hard to maintain equivalence ratio at 0.51 at 42% bustion duration for CDC was much longer than that for DFC mainly
EGR condition considering the limitation of the air handling system due to the multiple injection strategy and the more fraction of
of real engines, so the equivalence ratio was set at 0.81 and intake diffusion combustion. When the EGR rate increased to 42%,
pressure was 1.90 bar at the EGR rate of 42%. Triple injection equivalence ratio, intake temperature and intake pressure
strategy was adopted for CDC to mitigate high peak pressure rise increased, as illustrated in Table 5. For CDC, where diffusion com-
rate (PPRR), while single injection strategy was used for DFC. bustion was dominant, the combustion duration increased by 6 CA
Herein, the minimum injection pressure for CDC was 600 bar, approximately for the given injection pressures due to the EGR
because further reduction in injection pressure would significantly dilution effect. However, it seems that DFC was much less sensitive
to EGR rate since only slight change in combustion duration was
observed. This can be attributed to the combustion process. For
Table 4
DFC, natural gas was well premixed, the increased equivalence ratio
Uncertainties of measured parameters and calculated values.
and higher in-cylinder temperature after compression could partly
Parameter Uncertainty (%) counteract the EGR dilution effect.
Engine speed 0.5 Fig. 3(c) presents PPRR at different injection pressures for CDC
Crank angle 0.1 and DFC modes. Clearly, CDC has lower PPRR owing to the adoption
Gaseous emissions 2.0
of multiple injection strategy. PPRR was reduced obviously due to
AC 1.0
Cylinder pressure 1.5 the EGR addition and the resultant suppression of combustion. Low
Air flow meter 1.0 injection pressure was also beneficial for reducing PPRR because of
Fuel flow meter 1.0 the less fraction of pre-mixed combustion. For example, PPRR
IMEP 0.6 reduced 1.44 bar when the injection pressure changed from
ITE 0.8
1000 bar to 800 bar at CDC mode 0% EGR conditions.
 H. Song et al. / Energy 153 (2018) 248e255 251

Table 5
Experimental conditions.

Parameter EGR ¼ 0% EGR ¼ 42%

CDC Dual-fuel CDC Dual-fuel

Speed [rpm] 1000 1000 1000 1000

IMEP [bar] (approximately) 12 12 12 12
Substitution ratio [%] 0 90 0 90
Intake pressure [bar] 1.58 1.58 1.90 1.90
Intake temperature [K] 300.6 300.6 306.5 306.5
Equivalence ratio 0.51 0.51 0.81 0.81
Direct injection pressure [bar] 600e1000 200e1000 600e1000 200e1000
Pre-injection timing [ CA BTDC] 20 \ 20 \
Main-injection timing [ CA BTDC] 14 42 14 42
Post-injection timing [ CA BTDC] 6.3 \ 6.3 \

160 Injection Pressure Substitute Ratio: 0% 900 160 Injection Pressure Substitute Ratio: 0% 900
1000 bar: EGR Rate: 0% 1000 bar: EGR Rate: 42%
800 bar: 800 bar:
600 bar: 600 bar:
Cylinder Pressure (bar)

Cylinder Pressure (bar)

120 120
600 600
AHRR (J/deg)

AHRR (J/deg)
80 80

300 300
40 40

0 0 0 0

-30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30

Crank Angle (deg ATDC) Crank Angle (deg ATDC)

a) CDC at 0% EGR rate b) CDC at 42% EGR rate

200 1500 200 1500

Injection Pressure Substitute Ratio: 90% Injection Pressure Substitute Ratio: 90%
1000 bar: EGR Rate: 0% 1000 bar: EGR Rate: 42%
160 1200 160 1200
800 bar: 800 bar:
600 bar: 600 bar:
Cylinder Pressure (bar)

Cylinder Pressure (bar)

400 bar: 400 bar:

120 900 120 900
200 bar: 200 bar:
AHRR (J/deg)

AHRR (J/deg)

80 600 80 600

40 300 40 300

0 0 0 0
-30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30
Crank Angle (deg ATDC) Crank Angle (deg ATDC)

c) DFC at 0% EGR rate d) DFC at 42% EGR rate

Fig. 2. Impact of injection pressure on cylinder pressure and AHRR for CDC and DFC at different EGR rate conditions (dash lines represent AHRR).

The results of CO and THC emissions are shown in Fig. 4(a) and combustion efficiency at low equivalence ratio conditions. There-
Fig. 4(b), respectively. At the CDC mode, diffusion combustion fore, extremely low CO and THC emissions were observed at 0% EGR
dominated the whole combustion process, which resulted in high rate. When EGR rate increased, both combustion temperature and
252 H. Song et al. / Energy 153 (2018) 248e255

15 Diesel Dual-fuel 20
Diesel Dual-fuel Diesel Dual-fuel
50 EGR= 0%
EGR= 0% EGR= 0%

Comb. Duration(deg CA)

EGR=42%
EGR=42% EGR=42%
10 15

PPRR(bar/deg CA)
CA50(deg ATDC)

40

5 10
30

0 5
20
-5 0
200 400 600 800 1000 200 400 600 800 1000 200 400 600 800 1000
Injection Pressure(bar) Injection Pressure(bar) Injection Pressure(bar)

a) CA50 b) Combustion duration c) PPRR

Fig. 3. Impact of injection pressure on CA50, combustion duration and PPRR for CDC and DFC at different EGR rate conditions.

40 20
Diesel Dual-fuel Diesel Dual-fuel
EGR= 0% 15 EGR= 0%
EGR=42% EGR=42%
30
10
THC(g/kWh)
CO(g/kWh)

5
20
0.8

10
0.4

0 0.0
200 400 600 800 1000 200 400 600 800 1000
Injection Pressure(bar) Injection Pressure(bar)

a) CO b) THC

Fig. 4. Impact of injection pressure on CO and THC emissions for CDC and DFC at different EGR rate conditions.

Diesel Dual-fuel 40
12 Diesel Dual-fuel
EGR= 0%
EGR=42% 30 EGR= 0%
8 EGR=42%
20
4
NOx(g/kWh)

10
AC(1/m)

0.08
3
Europe V
0.04 2
1 Europe VI
0.00 0
200 400 600 800 1000 200 400 600 800 1000
Injection Pressure(bar) Injection Pressure(bar)

a) Soot b) NOx

Fig. 5. Impact of injection pressure on soot and NOx emissions for CDC and DFC at different EGR rate conditions.
 H. Song et al. / Energy 153 (2018) 248e255 253

oxygen fraction decreased, leading to the increase in CO and THC Fig. 5(a) shows the results of soot emissions. Soot emissions
emissions. Higher injection pressure resulted in longer spray generally followed similar trend as CO emissions at different con-
penetration length, better fuel atomization and faster droplet ditions. It could be concluded from 4-T map [25] that soot forma-
evaporation, which could improve fuel-air mixture quality and tion mostly occurred at high equivalence ratio conditions. Oxygen
counteract the effects of lower oxygen fraction and combustion fraction reduced significantly with EGR addition, which leaded to
temperature to some extent. Thus, lower CO and THC emissions higher equivalence ratio. Increasing injection pressure could pro-
were obtained under higher injection pressures. At the DFC mode, mote fuel atomization and reduce local equivalence ratio during the
THC emissions were significantly higher than CDC mode. This is combustion process, which leaded to lower soot emissions. At 0%
reasonable because the relatively lean pre-mixed mixture of nat- EGR conditions, low equivalence ratio and high combustion tem-
ural gas and air decreases fuel combustion efficiency and part of the perature resulted in extremely low soot emissions for both CDC and
natural gas cannot be fully oxidized [17]. In the cases without EGR, DFC. At the CDC mode, soot emissions were much higher at 42%
higher CO emissions were observed compared to CDC without EGR. EGR, especially at the low injection pressure. At the DFC mode, soot
However, DFC was not sensitive to EGR rate or injection pressure emissions kept at a low level as the EGR rate and injection pressure
and has comparable CO emissions with CDC without EGR, much varied. As shown Fig. 5(b), NOx emissions were higher than Europe
lower than CDC with EGR. V standard (2.0 g/kW$h) at both CDC and DFC modes when EGR
rate was 0%. The utilization of EGR could decrease oxygen fraction
and bulk temperature and result in lower NOx emissions. Thus,
NOx emissions were even lower than Europe VI standard (0.4 g/
kW$h) at all experiment conditions when 42% EGR rate was used.
Fig. 6 shows the trade-off between soot and NOx emissions. When
EGR rate increased, the significantly decreased NOx emissions was
expected. For CDC mode, soot emissions significantly increased as
illustrated by the NOx-soot trade-off. For DFC mode, however, no
obvious trade-off between NOx and soot was observed. Soot
emissions still kept a low level even with 200 bar injection pressure
at the increased EGR rate. Thus, natural gas DFC could break the
NOx-soot trade-off, which was usually observed in CDC. Low soot
and NOx emissions can be achieved simultaneously even at the
injection pressure of low to 200 bar.
Fig. 7 shows the PM size distributions. Overall, the total particle
number (PN) emissions followed the same trend as soot emissions.
The nucleation mode PM dominated for both CDC and DFC modes
when EGR rate was 0%. For dual-fuel mode, no obvious variation in
PM size distribution was observed at the two different EGR con-
ditions. PM emissions kept at a low level even at 200 bar injection
pressure. For CDC mode, both PM size and PN increased and the
accumulation mode PM became dominant at higher EGR rate. Both
Fig. 6. The trade-off between NOx and soot emissions. PM and PN emissions became even worse at lower injection

a) EGR=0% b) EGR=42%
Fig. 7. Impact of injection pressure on PM size distributions for CDC and DFC at different EGR rate conditions.
254 H. Song et al. / Energy 153 (2018) 248e255

56 were lower than Europe VI standard (0.4 g/kW$h) at 42%

Diesel Dual-fuel EGR. The DFC mode did not exhibit apparent trade-off be-
EGR= 0% tween soot and NOx emissions, which was clearly observed
EGR=42% at CDC mode. Overall, EGR decreased ITE. However, the ITE of
52 DFC was less sensitive to EGR compared to CDC.

The main challenge for LPDI strategy is how to keep PM emis-

48 sions at a low level under poor fuel atomization conditions.
ITE(%)

Reducing direct injection fuel amount and increasing ignition delay

can significantly reduce the requirement of fuel atomization qual-
44 ity. Natural gas DFC could achieve low PM emissions and high ef-
ficiency by a small amount of direct injection fuel. Advanced
injection timing can greatly increase ignition delay, which is usually
used for LTC concept. The present study has demonstrated the
40 possibility of utilizing LPDI strategy at natural gas DFC in
conjunction with LTC concept. In the future, more research will be
200 400 600 800 1000 conducted, including using alternative fuels with better atomiza-
tion properties rather than diesel, optimization of control param-
Injection Pressure(bar) eters and engine design.
Fig. 8. Impact of injection pressure on ITE for CDC and DFC at different EGR rate
conditions. Acknowledgements

This work is sponsored by the Project of National Natural Sci-

pressure. ence Foundation of China under the Grant No. 91541206.
The results of indicate thermal efficiency (ITE) are shown in
Fig. 8. Without EGR, DFC achieved the similar ITE as CDC. When EGR
rate increased, combustion duration, CO and soot emissions References
significantly increased at CDC, which contributed to the lower ITE.
[1] Gnanasekaran S, Saravanan N, Ilangkumaran M. Influence of injection timing
For the DFC mode, ITE decreased mainly due to the slower heat on performance, emission and combustion characteristics of a DI diesel engine
release rate considering that CO, THC and soot emissions hardly running on fish oil biodiesel. Energy 2016;116:1218e29.
increased with increasing EGR. Higher ITE was observed at DFC, [2] Qiu T, Li X, Liang H, et al. A method for estimating the temperature down-
stream of the SCR (selective catalytic reduction) catalyst in diesel engines.
indicated the better EGR tolerance than CDC. Taking 600 bar in- Energy 2014;68:311e7.
jection pressure as an example, ITE decreased 17.0% at CDC mode [3] Liu H, Wang Z, Zhang J, et al. Study on combustion and emission character-
with EGR increasing from 0% to 42%, while 8.7% reduction of ITE istics of Polyoxymethylene Dimethyl Ethers/diesel blends in light-duty and
heavy-duty diesel engines. Applied energy 2017;185:1393e402.
was observed at DFC mode. ITE decreased slightly as injection [4] Verschaeren R, Schaepdryver W, Serruys T, et al. Experimental study of NO x
pressure decreased mainly due to the longer combustion duration reduction on a medium speed heavy duty diesel engine by the application of
and the resultant loss in heat transfer, as shown in Fig. 3(b). EGR (exhaust gas recirculation) and Miller timing. Energy 2014;76:614e21.
[5] Yao M, Zheng Z, Liu H. Progress and recent trends in homogeneous charge
Considering the quite low PPRR at 42% EGR, as shown in Fig. 3(c), compression ignition (HCCI) engines. Prog Energy Combust Sci 2009;35(5):
the engine ITE of DFC can be further improved by increasing engine 398e437.
swirl ratio to promote the combustion process. [6] Jia M, Li Y, Xie M, et al. Numerical evaluation of the potential of late intake
valve closing strategy for diesel PCCI (premixed charge compression ignition)
engine in a wide speed and load range. Energy 2013;51:203e15.
4. Conclusions [7] Splitter D, Reitz RD, Hanson R. High efficiency, low emissions RCCI combustion
by use of a fuel additive. SAE Int J Fuel Lubr 2010;3:742e56 (2010-01-2167).
[8] How HG, Masjuki HH, Kalam MA, et al. An investigation of the engine per-
In the present work, the feasibility of LPDI strategy in natural gas
formance, emissions and combustion characteristics of coconut biodiesel in a
DFC was investigated and the comparison with CDC at different high-pressure common-rail diesel engine. Energy 2014;69:749e59.
injection pressure and EGR rate conditions were carefully analyzed [9] Payri R, Salvador FJ, Gimeno J, et al. Influence of injector technology on in-
in terms of combustion and emission characteristics. The conclu- jection and combustion developmentePart 1: hydraulic characterization. Appl
Energy 2011;88(4):1068e74.
sions can be summarized as follows: [10] The MARTEC Group, Inc. Variable costs of fuel economy technologies, Pre-
pared for the alliance of automobile manufacturers, June 1, 2008; amended
(1) Natural gas DFC can achieve low soot and NOx emissions December 10, 2008.
[11] Liu J, Yao A, Yao C. Effects of diesel injection pressure on the performance and
even at the injection pressure of 200 bar, and thus the utili- emissions of a HD common-rail diesel engine fueled with diesel/methanol
zation of GDI fuel supply systems instead of CRI systems is an dual fuel. Fuel 2015;140:192e200.
option for natural gas/diesel dual fuel combustion, which [12] Walker NR, Dempsey AB, Andrie MJ, et al. Use of low-pressure direct-injection
for reactivity controlled compression ignition (RCCI) light-duty engine oper-
could significantly reduce the system cost without the pen- ation. SAE Int J Engines 2013;6:1222e37 (2013-01-1605).
alty of emissions. [13] Chang Z, Yu C, Zhang H, et al. Experimental investigation of homogeneous
(2) For CDC mode, CO and THC emissions increased obviously charge induced ignition (HCII) with low-pressure injection to reduce Pm
emissions in a heavy-duty engine. SAE Int J Engines 2016;9:1559e70 (2016-
with increasing EGR rate, which became much worse at LPDI 01-0775).
strategy. For DFC mode, higher CO and THC emissions were [14] Ryu K. Effects of pilot injection pressure on the combustion and emissions
observed in comparison with CDC when EGR rate was 0%.CO characteristics in a diesel engine using biodieseleCNG dual fuel. Energy
Convers Manag 2013;76:506e16.
and THC emissions of DFC were not sensitive to EGR rate or
[15] Sun L, Zhou L, Zeng K. Experimental investigation of cycle-by-cycle variations
injection pressure. in a natural gas/diesel dual fuel engine with EGR. SAE Technical Paper, 2013-
(3) For DFC mode, slight variation in PM size distribution and 01-0853.
total PN was observed at all injection pressure and EGR [16] Ogawa H, Zhao P, Kato T, et al. Improvement of combustion and emissions in a
dual fuel compression ignition engine with natural gas as the main fuel. SAE
conditions. For CDC mode, higher EGR rate and lower injec- Technical Paper, 2015-01-0863.
tion pressure increased PM and PN emissions. NOx emissions [17] Song H, Liu C, Li F, et al. A comparative study of using diesel and PODEn as
 H. Song et al. / Energy 153 (2018) 248e255 255

pilot fuels for natural gas dual-fuel combustion. Fuel 2017;188:418e26. optimization of pollutant emissions and combustion noise. Energy 2017;134:
[18] Guerry ES, Raihan MS, Srinivasan KK, et al. Injection timing effects on partially 420e37.
premixed dieselemethane dual fuel low temperature combustion. Appl En- [22] Coleman HW, Steele WG. Experimentation, validation, and uncertainty anal-
ergy 2016;162:99e113. ysis for engineers. John Wiley & Sons; 2009.
[19] McTaggart-Cowan G, Mann K, Huang J, et al. Direct injection of natural gas at [23] Li Y, Xu H. Experimental study of temporal evolution of initial stage diesel
up to 600 bar in a pilot-ignited heavy-duty engine. SAE Int J Engines 2015;8 spray under varied conditions. Fuel 2016;171:44e53.
(2015-01-0865). [24] Wang Z, Li Y, Wang C, et al. Experimental study on primary breakup of diesel
[20] Paul A, Panua RS, Debroy D, et al. An experimental study of the performance, spray under cold start conditions. Fuel 2016;183:617e26.
combustion and emission characteristics of a CI engine under dual fuel mode [25] Dempsey AB, Walker NR, Gingrich E, et al. Comparison of low temperature
using CNG and oxygenated pilot fuel blends. Energy 2015;86:560e73. combustion strategies for advanced compression ignition engines with a focus
[21] d’Ambrosio S, Ferrari A. Boot injection dynamics and parametrical analysis of on controllability. Combust Sci Technol 2014;186(2):210e41.
boot shaped injections in low-temperature combustion diesel engines for the