Applied Energy 165 (2016) 166–182

Contents lists available at ScienceDirect

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

Experimental analysis of ethanol dual-fuel combustion in a heavy-duty

diesel engine: An optimisation at low load
Vinícius B. Pedrozo ⇑, Ian May, Macklini Dalla Nora, Alasdair Cairns, Hua Zhao
Brunel University London, Centre for Advanced Powertrain and Fuels Research (CAPF), Kingston Lane, Uxbridge, Middlesex UB8 3PH, United Kingdom

h i g h l i g h t s

Dual-fuel combustion offers promising results on a stock heavy-duty diesel engine.

The use of split diesel injections extends the benefits of the dual-fuel mode.

Ethanol–diesel dual-fuel combustion results in high indicated efficiencies.

NOx and soot emissions are significantly reduced.

Combustion efficiency reaches 98% with an ethanol energy ratio of 53%.

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

Article history: Conventional diesel combustion produces harmful exhaust emissions which adversely affect the air qual-
Received 10 August 2015 ity if not controlled by in-cylinder measures and exhaust aftertreatment systems. Dual-fuel combustion
Received in revised form 7 November 2015 can potentially reduce the formation of nitrogen oxides (NOx) and soot which are characteristic of diesel
Accepted 14 December 2015
diffusion flame. The in-cylinder blending of different fuels to control the charge reactivity allows for
lower local equivalence ratios and temperatures. The use of ethanol, an oxygenated biofuel with high
knock resistance and high latent heat of vaporisation, increases the reactivity gradient. In addition,
Keywords:
renewable biofuels can provide a sustainable alternative to petroleum-based fuels as well as reduce
Dual-fuel combustion
Ethanol
greenhouse gas emissions. However, ethanol–diesel dual-fuel combustion suffers from poor engine effi-
Split diesel injections ciency at low load due to incomplete combustion. Therefore, experimental studies were carried out at
Engine-out emissions 1200 rpm and 0.615 MPa indicated mean effective pressure on a heavy-duty diesel engine. Fuel delivery
Combustion losses was in the form of port fuel injection of ethanol and common rail direct injection of diesel. The objective
Low load was to improve combustion efficiency, maximise ethanol substitution, and minimise NOx and soot emis-
sions. Ethanol energy fractions up to 69% were explored in conjunction with the effect of different diesel
injection strategies on combustion, emissions, and efficiency. Optimisation tests were performed for the
optimum fuelling and diesel injection strategy. The resulting effects of exhaust gas recirculation, intake
air pressure, and rail pressure were investigated. The optimised combustion of ethanol ignited by split
diesel injections resulted in higher net indicated efficiency when compared to diesel-only operation.
For the best emissions case, NOx and soot emissions were reduced by 65% and 29%, respectively.
Aftertreatment requirements that are generally associated with cost and fuel economy penalties can
be minimised. Combustion efficiency of 98% was achieved at the expense of higher NOx emissions.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction capability, reliability, as well as superior fuel conversion efficiency

Heavy-duty (HD) diesel engines have been widely utilised in on
and off-road transportation sectors due to their high torque
⇑ Corresponding author at: Brunel University London, Centre for Advanced
Powertrain and Fuels Research (CAPF), Kingston Lane, Uxbridge, Middlesex UB8
3PH, United Kingdom. Tel.: +44 7473361328; fax: +44 1895266698.
E-mail addresses: vinicius.pedrozo@brunel.ac.uk, pedrozo.vinicius@gmail.com
(V.B. Pedrozo).
[1]. However, conventional diesel combustion incurs a wide range
of local in-cylinder equivalence ratios and temperatures which can
result in NOx and soot formation [2]. These emissions can
adversely affect the air quality if not controlled by exhaust
aftertreatment technologies and in-cylinder measures. Several
combustion concepts have been developed, arising from costly
aftertreatment systems and strict fuel efficiency and emissions
regulations [3]. In addition, according to economic growth

http://dx.doi.org/10.1016/j.apenergy.2015.12.052
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182
projections, it is predicted an increase in the demand for petroleum
and other energy sources by more than 30% from 2010 to 2040 [4].
This may result in elevated prices for liquid fuels as well as
compromise their cost competitiveness, opening opportunities
for improved sustainability and greenhouse gas (GHG) emissions
reduction via biofuels [5].
The alternative combustion technologies are generally centred
on improved fuel atomisation and mixture preparation, lower local
equivalence ratios, reduced peak in-cylinder temperatures, and
faster burn rates. This is usually referred to Low Temperature Com-
bustion (LTC) [3]. Among the combustion strategies proposed is
Homogeneous Charge Compression Ignition (HCCI). This is charac-
terised by early fuel injections promoting a fully premixed charge,
long ignition delays, and short combustion durations. However, the
lack of direct control of ignition timing and combustion phasing,
particularly under transient conditions, is still the major drawback.
It also exhibits elevated combustion losses, combustion noise, and
sensitivity to temperature [6–8]. In comparison, some slightly
more heterogeneous combustion concepts have been developed.
Premixed Charge Compression Ignition (PCCI) [9–12], Partially Pre-
mixed Charge Compression Ignition (PPCI) [13], Modulated Kinet-
ics (MK) [14], and Uniform Bulky Combustion System (UNIBUS)
[15] name a few. These allow a higher degree of combustion phas-
ing control at low and medium loads while maintaining low soot
and NOx emissions. However, these less premixed combustion
modes tend to suffer from lower indicated efficiency, increased
unburnt hydrocarbons (HC) and carbon monoxide (CO) emissions,
and limited load range due to high exhaust gas recirculation (EGR)
and boost requirements.
Gasoline Direct Injection Compression Ignition (GDCI) [16,17]
and Partially Premixed Combustion (PPC) [18–20] are some alter-
natives to diesel LTC. They expand the high efficiency window
and achieve very low NOx emissions operating up to full load with
moderate-high EGR rates. As these concepts utilise gasoline, they
do not reduce the dependence on liquid fossil fuels. They also
require engine hardware modifications such as the piston and
injection system, and ignition or lubricant improvers, depending
on the fuel selected. Some drawbacks regarding soot levels at
higher loads, due to low air–fuel ratio, accompanied with
significant CO and HC emissions at low loads are also reported.
Recent PPC studies with renewable fuels, including ethanol, have
demonstrated high thermal efficiency and further soot reductions
[21–23]. However, high acoustic noise and elevated peak heat
release rates have been experienced due to a fast premixed com-
bustion, requiring lower intake air pressures and larger amounts
of EGR, which reduce combustion efficiency [24].
Dual-fuel (DF) combustion, such as Reactivity Controlled Com-
pression Ignition (RCCI) [25–27], have been developed to overcome
the majority of the previously mentioned issues. The concept uses
different fuels to control the in-cylinder reactivity gradient while
achieving a wide operating range with near zero levels of NOx
and soot, acceptable pressure rise rate (PRR), and very high indi-
cated efficiency [28]. The primary method of fuel delivery is the
port fuel injection of a low reactivity fuel (i.e. gasoline, alcohol,
propane, natural gas, etc.) to create a well-mixed charge of fuel–
air-EGR. The high reactivity fuel (i.e. diesel) serves as the ignition
source and is directly injected into the combustion chamber. How-
ever, RCCI is sensitive to variations in the intake air temperature
and pressure. This is expected as the combustion is sufficiently
premixed and governed by chemical kinetics [25]. Furthermore,
the combustion phasing is generally controlled by varying the fuel
reactivity (i.e. substitution ratio), which might not be the optimum
at certain engine loads.
Ethanol is attractive as a low reactivity fuel because it can be
produced from biomass and can offset the demand for
petroleum-based fuels in internal combustion engines [29]. The
167
elevated knock resistance and latent heat of vaporisation of the
ethanol allow the use in high compression ratio and highly boosted
engines [30]. Moreover, early dual-fuel results obtained from an
optical engine showed that ethanol can suppress soot formation
in high temperature regions of the conventional diesel combustion
chamber [31]. Recent experimental analyses with ethanol–diesel
DF combustion demonstrated noticeable NOx reductions at engine
loads above 0.8 or 1.0 MPa net indicated mean effective pressure
(IMEP) [32–36].
Asad et al. [37] investigated the load range of ethanol–diesel
low temperature combustion using a single cylinder light-duty
engine. The concept, named Premixed Pilot Assisted Combustion
(PPAC), utilised high EGR levels and single diesel injections near
firing top dead centre (TDC). The main challenge encountered
was the elevated levels of unburnt HC and CO emissions at low
loads, as observed by other ethanol dual-fuel combustion studies
[38,39]. Asad et al. [37] attributed these losses to the resistance
of ethanol to auto-ignition and proposed an alternative combus-
tion strategy to enable clean combustion and higher efficiencies
at these specific conditions. From idle to low loads, the engine
would operate under conventional diesel combustion, utilising
high levels of EGR and boost combined with retarded injections
and elevated rail pressures. After the engine reaches a certain load,
the combustion would switch to ethanol dual-fuel combustion.
However, high levels of EGR might not be feasible and would place
greater demand on the boosting system to maintain the required
equivalence ratio.
Sarjovaara et al. [40] studied the effect of diesel injection
parameters on ethanol dual-fuel combustion using a modified
six-cylinder diesel engine with a compression ratio of 14.2:1. The
use of small pilot injections corresponding to approximately 10–
20% of the total diesel fuel injected helped maintaining acceptable
PRR levels while running with high ethanol percentages. However,
the maximum ethanol substitution ratio was limited to only 39% at
25% load, and exhaust gas emissions were neglected.
Considering the previous works, an experimental study was
carried out on a single cylinder HD engine. The combustion system
remained stock, with a re-entrant piston bowl design and a geo-
metric compression ratio of 16.8:1. Ethanol was port fuel injected
while diesel fuel was directly injected into the cylinder. The objec-
tive was to reduce combustion losses and maximise the use of
ethanol while maintaining low levels of NOx and soot emissions.
The diesel injection strategy tested uses a pre-injection to adjust
mixture flammability and reduce PRR, and one injection around
TDC to maintain combustion control. This concept is slightly differ-
ent from conventional dual-fuel using a single diesel injection near
TDC for ignition [33–37,41]. The strategy also differs from the early
diesel injections utilised in RCCI combustion [25,26,28].
The experiments were performed at 1200 rpm and 0.615 MPa
IMEP, with varying ethanol energy fractions up to 69%. The impact
of different diesel injection strategies on combustion, emissions,
and efficiency were explored. Pre-injections corresponding to up
to 60% of the total diesel fuel injected were evaluated without
EGR. Subsequent investigation of the pre-injection timing and
quantity was performed for the optimum fuelling and injection
strategy using an EGR rate of 25%. Finally, the effect of higher
intake air pressure and diesel injection pressure were explored.
The best dual-fuel results were compared against diesel-only
operation.
2. Experimental setup
The experiments were carried out on a single cylinder HD diesel
engine equipped with a high pressure common rail diesel injection
system, representing the engine of a modern heavy goods vehicle
168 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

(i.e. truck). The test cell layout and main engine specifications are Table 1
depicted in Fig. 1 and Table 1, respectively. Measurement device Single cylinder HD diesel engine specifications.

specifications are shown in the Appendix A. Two large-volume Parameter Value

surge tanks were installed to damp out pressure fluctuations in Bore 129 mm
the intake and exhaust manifolds. Fresh intake air was supplied Stroke 155 mm
to the engine via an external compressor system with closed loop Swept volume 2026 cm3
control for the boost pressure. An intake throttle provided fine con- Geometric compression ratio 16.8:1
Maximum in-cylinder pressure 18 MPa
trol over the intake manifold air pressure. The fresh air flow rate Piston type Re-entrant bowl
was measured with a thermal mass flow meter. High-pressure loop Number of valves 4
cooled external EGR was supplied to the engine via an EGR valve Diesel injection system Common rail, inj. pressure of 50–220 MPa,
and an electronically controlled exhaust back pressure valve centrally mounted diesel injector, 8 holes
Ethanol injection system PFI peak-and-hold Marelli IWP069,
located downstream of the exhaust surge tank.
included spray angle of 15°
Auxiliary equipment such as the high-pressure diesel pump
(HPP) and the engine coolant and oil pumps are not coupled to
the engine but driven by separate electric motors. Coolant and
oil temperatures were kept within 353 ± 5 K. Oil pressure was set Table 2
Fuel properties.
to 350 ± 10 kPa throughout the experiments. An independent
low-pressure system supplied diesel to the common rail injection Characteristic Diesel Ethanol
system. Two Coriolis flow meters were used to measure the diesel Product Gasoil (Ultra Low Sulphur) Ethyl alcohol
flow rate (m _ diesel ) by considering the total fuel supplied to and from Density at 293 K 827 kg/m3 789 kg/m3
the injector and HPP. Ethanol was injected into the intake port Cetane number 45 –
RON [42] – 107
through a high flow-rate peak-and-hold port fuel injector (PFI). It
Alcohol content NA 99.1–99.5% (v/v)
was mounted into the intake air manifold so that the spray was Water content <0.2 g/kg <1.14 (w/w)
directed towards the back of the intake valves, located approxi- Boiling point/range 443–643 K 351 K
mately 0.3 m downstream. Heat of vaporisation [42] 270 kJ/kg 840 kJ/kg
An in-house injector driver controlled the PFI pulse width, Carbon content 86.6% 52.1%
Hydrogen content 13.2% 13.1%
adjusted according to the desired ethanol substitution ratio. The Oxygen content 0.2% 34.8%
ethanol start of injection (SOI) was set to the firing TDC to max- LHV 42.9 MJ/kg 26.9 MJ/kg
imise the time for air–fuel mixture preparation before the intake
valve opening event. The ethanol mass flow rate (m _ ethanol ) was
obtained from the injector calibration curve. Ethanol injection emissions [43] and confirmed by the air and fuel flow rates. The
pressure was continuously monitored by a pressure transducer, lower heating value of the DF combustion mode (LHVDF) was calcu-
so that a constant delta pressure of 300 ± 10 kPa could be main- lated using Eq. (1).
tained across the injector. A heat exchanger held the fuel temper- _ ethanol LHVethanol þ m
m _ diesel LHVdiesel
ature constant at 293 ± 5 K. The relevant properties of the fuel used LHVDF ¼ ð1Þ
_ ethanol þ m
ðm _ diesel Þ
in this work are listed in Table 2.
The stoichiometric air/fuel ratio was determined by the conser- The ethanol substitution ratio (EF) was calculated on an energy
vation of mass of each chemical element in the reactants [42]. The input basis, defined as the ratio of the energy content of the etha-
global equivalence ratio was calculated based on the engine-out nol to the total fuel energy (Eq. (2)).

Fig. 1. Schematic diagram of the engine experimental setup.

 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182 169

_ ethanol LHVethanol
m engine operation condition is close to operation points #3 and
EF ¼ ð2Þ
_ ethanol LHVethanol þ m
m _ diesel LHVdiesel #8 of the World Harmonized Stationary Cycle (WHSC) and #7 of
the former European Stationary Cycle (ESC13) for HD engines.
The in-cylinder pressure was measured by a piezoelectric pres-
Fig. 2 shows where the test point (white circle with black dots pat-
sure sensor. Intake and exhaust pressures were measured by two
tern) is located over an estimated speed and load map of a HD die-
water cooled piezoresistive absolute pressure sensors. The intake
sel engine. The WHSC and ESC13 test cycle points are also
valve lift profile was obtained by measuring the displacement of
displayed. The bigger the circle, the higher is the relative weight
the valve spring retainer with a displacement sensor. Tempera-
over the test cycle. The selected test point is located in a high res-
tures and pressures at relevant locations were measured by K-
idency area of a typical HD diesel engine drive cycle. It also repre-
type thermocouples and pressure gauges, respectively. Two
sents an engine load where dual-fuel combustion is adversely
National Instruments data acquisition (DAQ) cards were used to
affected by elevated CO and unburnt HC emissions. The main goal
acquire the signals from the measurement device. While a high
of this study was to mitigate combustion losses and improve effi-
speed DAQ card received the crank angle resolved data synchro-
ciency while achieving NOx and soot levels close to Euro VI legis-
nised with an optical encoder of 0.25 crank angle degrees (CAD)
lation emissions limits (0.40 and 0.01 g/kW h, respectively)
resolution, a lower speed DAQ card acquired the low frequency
utilising low levels of intake air pressure and EGR.
engine operation conditions. These data were collected and dis-
The engine is equipped with a prototype variable valve actua-
played live by an in-house DAQ program. The pumping work was
tion system (VVA). Variable intake valve closing timing (IVC) and
included in all the calculations, such as IMEP and net indicated effi-
the resulting effective compression ratio (ECR) can be selected dur-
ciency (g_ind). The apparent net heat release rate (HRR), denoted
ing the engine operation. The intake valve opening and closing tim-
by (dQn/dt), was calculated using the following well-known Eq. (3):
ings were set at 367 CAD and 150 CAD after firing top dead centre
dQ n c dV 1 dp (ATDC), respectively, at 0.5 mm valve lift. This valve timing pro-
¼ p þ V ð3Þ
dt c  1 dt c  1 dt vided an ECR of approximately 16.0:1. The average PRR and COV_-
IMEP limits were set to 2 MPa/CAD and 5%, respectively.
where c is the ratio of specific heats, t is time, and V and p stand for
in-cylinder volume and pressure, respectively. Since the absolute
value of heat released is not as important to this study as the bulk 4. Results and discussion
shape of the curve with respect to crank angle, a c of 1.33 was
assumed. Combustion phasing (CA50) was determined by the crank 4.1. Conventional diesel combustion baseline
angle position of 50% mass fraction burnt (MFB). Ignition delay was
defined as the period of time between the diesel start of injection The primary objective of these tests was to obtain the best indi-
and start of combustion (SOC), set to 0.3% MFB point of the average cated specific fuel consumption (ISFC)/NOx/soot trade-off in
cycle. Cycle-to-cycle variability was measured by the coefficient of diesel-only operation by means of moderate amounts of EGR, ele-
variation of the IMEP (COV_IMEP), defined as the ratio of the stan- vated injection pressures, and optimised SOI. The conventional die-
dard deviation in IMEP and the mean IMEP over the sampled cycles. sel baseline was performed at four different EGR rates of 0%, 10%,
For the sake of simplification, the average in-cylinder gas tempera- 21%, and 25%. EGR temperature varied from 296 to 354 K as its
ture was calculated by applying the ideal gas model, considering ratio increased. Intake manifold air temperature also rose from
each species in the mixture. 292 K with no EGR to 306 K at 25% EGR. Two intake air pressures
Exhaust emissions were measured by a Horiba MEXA-7170 of 103 and 125 ± 1 kPa were included as a reference to the DF
DEGR emission analyser equipped with a heated line and a high experiments. A 10 kPa difference between intake air and exhaust
pressure module, allowing high-pressure sampling. The EGR rate gas back pressure was applied throughout the tests to maintain
was calculated by the ratio of intake and exhaust CO2 concentra- consistent pumping losses. The maximum COV_IMEP observed
tions measured by the same analyser. According to [44–46], deter- during these tests was 1.8%.
mining the actual hydrocarbons emissions measured by the flame Rail pressure (RP) was increased from 70 to 125 MPa as larger
ionisation detector (FID) needs to be calibrated for oxygenated percentages of EGR were added. A single injection strategy could
compounds such as ethanol. This is due to relative insensitivity only be applied in some cases in order to keep the PRR under the
of the equipment towards alcohols and aldehydes. Therefore, the limit of 2 MPa/CAD. A small pilot injection of approximately
FID response to ethanol was corrected by the method developed 3 mm3, with a dwell timing of 1 ms (7.2 CAD) and a split ratio of
in [45] with an updated factor of 0.68 [46]. This correction uses a
second order polynomial and the volumetric ethanol content as
an input. Smoke was measured by an AVL 415SE Smoke Meter.
The results were converted from FSN to mg/m3, according to
[47]. The calculation of specific exhaust gas emissions was based
on the UN Regulation number 49, with NOx and CO emissions cor-
rected to a wet basis [48]. Finally, combustion efficiency (gc) was
calculated using the following Eq. (4):
ðISCO LHV þ ISHC LHV Þ P
gc ¼ 1  _ CO DF i
_ diesel LHVdiesel Þ
ð4Þ
ðmethanol LHVethanol þ m
where ISCO and ISHC represent the net indicated specific emissions
of CO and unburnt HC in g/kW h, respectively, and Pi represents the
net indicated power.

3. Test conditions

The test point selected for this study was an engine speed of Fig. 2. The selected test point, and the WHSC and ESC13 test cycle points over an
1200 rpm and a load of 25%, equivalent to 0.615 MPa IMEP. This estimated HD diesel engine speed-load map.
170 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182
28/72, in average, was used to decrease the pressure rise rates. The
split ratio calculation was based on the ratio of the energising time
(ET) of each injection to the total energising time. The pilot injec-
tion resulted in shorter ignition delay (ID) periods (i.e. SOI_2 to
SOC), reducing the rate of premixed combustion represented by
the first peak heat release in Fig. 3. The reduced combustion noise
(PRR) was achieved at the expense of higher soot emissions, linked
to the formation of high equivalence ratio zones at intermediate
temperatures [1].
The sensitivity of the PRR, ISFC, ISNOx and ISSoot to different
intake air pressures, injection strategies, EGR rates, and combus-
tion phasing (CA50) are shown in Fig. 4. The legend makes refer-
ence to the intake pressure, EGR rate, RP, and split ratio. A higher
intake air pressure of 125 kPa improved fuel efficiency as leaner
mixtures reduced in-cylinder temperatures and subsequent heat
transfer losses [49]. Higher oxygen availability and higher injection
pressures reduced soot emissions by minimising fuel-rich combus-
tion and enhancing diesel atomisation and mixing. NOx production
decreased as more EGR was added due to lower combustion tem-
peratures, a result of the higher heat capacity of the charge and
lower oxygen concentration [50]. These are typical trade-offs of a
conventional diesel combustion system and directly related to
combustion temperature and local equivalence ratio.
Higher diesel fuel injection pressure and the ‘correct’ EGR level
resulted in an optimum NOx/soot trade-off without penalising fuel
consumption. This operation point will be used for future compar-
isons to the best DF combustion results. The selected diesel-only
calibration is circled on the curve denoted with ‘‘x” markers. It
was achieved at an intake air pressure of 125 kPa, a rail pressure
of 125 MPa, and 25% EGR, resulting in ISSoot and ISNOx emissions
of 0.018 and 2.01 g/kW h, respectively. A detailed summary of this
operating condition is given in Table 6 and will be discussed later
in the paper.
Fig. 3. In-cylinder pressure, diesel injection, and HRR curves of conventional diesel
combustion running without a pilot injection and with a split ratio of 26/74.
4.2. Determination of the optimum fuelling and diesel injection
strategy in DF combustion mode
In this section, the optimum ethanol substitution ratio and die-
sel injection strategy to ignite and efficiently burn the well-mixed
charge is identified. No external EGR was used in this initial phase
to reduce the complexity of the test. Intake air and exhaust back
pressures were held constant at 103 ± 1 kPa and 113 ± 1 kPa,
respectively. Intake manifold air temperature was maintained at
295 ± 3 K throughout this set of experiments. Upon using a single
diesel injection close to TDC, dual-fuel combustion had limited
operating range due to low fuel conversion efficiency and high
PRR [40,41], caused by fast heat release of the diesel fuel. Split die-
sel injections were then adopted to improve the mixture reactivity
stratification and enhance the ignition process.
In this test, the start of the second diesel injection (SOI_2) was a
result of the stipulated start of the first injection (SOI_1) and the
dwell timing between injections. The resulting SOI_2 remained
within 12 to 2 CAD ATDC. This differs from RCCI operation, where
the second injection is generally delivered near 35 CAD ATDC
[27]. The required energising time for the first injection (ET_1)
was set using the ECU’s application program. The energising time
of the second injection (ET_2) was automatically adjusted by the
engine speed governor. The estimate of the quantity injected dur-
ing ET_1 and ET_2 was based on the ratio of the energising time of
each injection to the total injection time (i.e. split ratio), as the die-
sel fuel mass injected at each energising time cannot be easily
determined.
The ethanol mass flow rates tested in this experimental analysis
were 34.5, 53.6, and 72.0 mg/inj. The lowest injection quantity was
limited by the minimum pulse width provided by the injector
drive. Considering this initial value represented approximately
40% on a mass input basis, the next two pulse widths were set at
around 60% and 80% of the total fuel mass injected. These fractions
changed as the indicated efficiency increased during the injection
strategy sweeps. A more detailed summary of the resultant ethanol
fractions for each fuelling and diesel injection strategy is depicted
in Table 3. The data were placed beside a similar diesel-only oper-
ating condition (i.e. not the optimum), used as a baseline in this
first section. The diesel injection pressure was held constant at
70 MPa.
The longest ET_1 of 0.90 ms could not be tested at the lowest
ethanol mass flow rate of 34.5 mg/inj because of rapid heat release.
At the highest ethanol substitution ratio (i.e. 72.0 mg/inj), an ET_1
of 0.43 ms was not applied as a short pre-injection did not allow
enough time for mixture preparation prior to the SOC, causing
knock. A longer dwell timing of 2 ms (⁄) was required with an
ET_1 of 0.59 ms to avoid elevated PRR. This particular injection
strategy was plotted with a label of 1 ms with the aim of simplify-
ing the analysis of the results. An ET_1 of 0.90 ms was removed by
the ECU at 72.0 mg/inj because ET_2 was too short to be
maintained.
Fig. 5 depicts the ignition delay from SOI_1 and SOI_2 to SOC for
those fuelling and diesel injection strategies showed in Table 3. The
legend makes reference to the ET_1 and dwell timing, while the
colours differentiate the fuelling strategy. It is observed that the
ignition delay from SOI_1 to SOC steadily rises for longer and more
advanced pre-injections. The increased amount of diesel injected
earlier in the cycle resulted in better mixture preparation, reducing
high reactivity zones within the piston bowl. The same behaviour
is seen when employing elevated ethanol fractions, due to lower
charge reactivity (i.e. higher RON) and lower bulk gas temperatures
introduced by the higher heat of vaporisation of the ethanol
[37,41]. While maintaining the same diesel injection strategy of
the diesel-only mode (ET_1 of 0.43 ms, dwell timing of 1 ms), the
utilisation of 34.5 and 53.6 mg of ethanol resulted in shorter
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182 171

3.0 202
Higher intake air
No pilot injection pressure,
EGR rate, and RP
198
PRR [MPa/CAD]

ISFC [g/kWh]
2.0

194

1.0
190

Higher RP

0.0 186
0 3 6 9 12 15 0 3 6 9 12
CA50 [CAD ADTC] ISNOx [g/kWh]

0.075

103 kPa, 0% EGR, RP 70 MPa, No Pilot

Higher intake air 103 kPa, 0% EGR, RP 70 MPa, 25/75
ISSoot [g/kWh]

0.050
pressure,
EGR rate, and RP 125 kPa, 0% EGR, RP 70 MPa, No Pilot

125 kPa, 0% EGR, RP 70 MPa, 26/74

0.025 125 kPa, 10% EGR, RP 90 MPa, 28/72

125 kPa, 21% EGR, RP 110 MPa, 28/72

125 kPa, 25% EGR, RP 125 MPa, 30/70

0.000
0 3 6 9 12
ISNOx [g/kWh]

Fig. 4. Conventional diesel combustion sensitivities of PRR, ISFC, ISNOx and ISSoot to different intake air pressures, injection strategies, EGR rates, and CA50.

Table 3
Fuelling and diesel injection strategies.

Parameter Unit Value

Ethanol flow rate mg/inj 0.0 34.5 53.6 72.0
Ethanol mass fraction – 0.00 0.41 0.43 0.44 0.58 0.61 0.63 0.65 0.77 0.78
Ethanol energy fraction – 0.00 0.31 0.32 0.33 0.47 0.50 0.52 0.54 0.67 0.69
LHVDF MJ/kg 42.9 36.3 36.1 35.9 33.6 33.1 32.8 32.5 30.6 30.4
ET_1 ms 0.43 0.43 0.59 0.76 0.43 0.59 0.76 0.90 0.59 0.76
Diesel inj. split ratio – 25/75 28/72 37/63 45/55 31/69 40/60 49/51 60/40 42/58 52/48
Dwell timing ms 1 1 1 3 1 1 3 5 2(⁄) 3

ignition delays than diesel-only operation. The reduced mixing
times might be a result of high reactivity caused by the greater
amount of diesel injected to compensate the combustion ineffi-
ciencies and keep the load constant. When the ignition delay
between the SOI_2 and the SOC was plotted, negative values were
observed for the more advanced and longer pre-injections (i.e. tri-
angles and squares). It depicts the fact that the SOI_2 took place
after combustion had already started. This was attributed to the
progressive auto-ignition of diesel fuel injected at SOI_1 and the
entrained air and ethanol in the piston bowl [26].
Fig. 6 shows that the combustion phasing was shifted linearly
towards TDC for a pre-injection taking place prior to 20 CAD
ATDC. However, no clear correlation between CA50 and SOI_1
could be observed as the pre-injections were advanced and the
energising times were lengthened. In some cases, a higher degree
of stratification in the piston bowl promoted by a retarded SOI_1
advanced the CA50 position. The auto-ignition of the premixed die-
sel was hindered by the lower reactivity of the ethanol-air charge
in the cylinder during the first injection and was more prone to
cyclic variations of flow and mixture motion. Despite the different
trend in CA50, the combustion duration (CA10–CA90) decreased as
the diesel injections were advanced as they allowed more mixing
time prior to combustion. For the same diesel injection strategy,
the combustion duration was shortened as more ethanol was
injected. This is a result of the faster combustion promoted by its
homogeneous distribution and flame propagation, generally lead-
ing to higher PRR. Combustion remained stable with COV_IMEP
in the range of 1.1–2.3% throughout the tests.
The results in Fig. 7 show that a lower NOx/soot emissions
trade-off can be achieved using DF combustion when compared
to equivalent conventional diesel combustion. Elevated levels of
soot with low ethanol fractions were associated with shorter igni-
tion delays and rapid rate of premixed combustion of the diesel
fuel. Higher ethanol fractions and earlier pre-injections helped
lowering soot emissions due to lower overall carbon to hydrogen
mass ratio and reduced diesel diffusion combustion [51]. Addition-
ally, the use of advanced and longer SOI_1 improved the mixture
preparation by creating multiple ignition sites, reducing charge
stratification and NOx emissions [52].
Fig. 8 depicts the combustion and net indicated efficiencies.
Indicated efficiency decreased for shorter and retarded pre-
injections due to reduced in-cylinder reactivity gradient and
172 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

Fig. 5. Ignition delay from SOI_1 and SOI_2 to SOC.

delayed combustion process towards the expansion stroke. Com-

bustion efficiency was associated with the amount of ethanol
injected. The reduced amount of diesel fuel (i.e. reduced in-
cylinder charge reactivity) inhibits the combustion progression,
resulting in increased unburnt HC and CO emissions levels [53].
Combustion modelling also showed that the majority of late cycle
unburnt HC is located in the piston-to-liner crevices, and that the
bulk of the CO resides near the liner and the crevice regions [54].
This phenomenon is exacerbated at elevated ethanol fractions
due to higher amount of fuel trapped in the crevices and squish
volumes of the conventional diesel combustion system, not Fig. 6. CA50, CA10–CA90 and PRR.
designed to operate on a premixed fuel [28].
The optimum DF operating mode was achieved using an etha-
nol mass flow rate of 53.6 mg/inj, with a SOI_1 between 40 and 0.24
0.43 ms, Dwell 1 ms Diesel
35 CAD ATDC and a split ratio of approximately 60/40. This fuel-
0.59 ms, Dwell 1 ms* 34.5 mg/inj
ling strategy is equivalent to an ethanol energy fraction of 54%. The 0.76 ms, Dwell 3 ms 53.6 mg/inj
diesel injection strategy selected allowed for appropriate reactivity 0.18
ISSoot [g/kWh]

0.90 ms, Dwell 5 ms 72.0 mg/inj

gradient. Combustion efficiency reached levels of a lower ethanol
substitution ratio. It is believed that higher intake air temperatures
0.12
could improve the ethanol evaporation process and thus combus-
tion efficiency at higher ethanol fractions [55]. A lower substitution
ratio (34.5 mg/inj) did not show advantages, worsening exhaust
0.06
emissions and net indicated efficiency.
Fig. 9 shows a comparison between the optimum DF
combustion strategy and an equivalent diesel-only case. Lower
0.00
in-cylinder pressure and bulk gas temperature were observed dur- 0 3 6 9 12
ing the compression stroke, prior to the SOC. This is mainly attrib- ISNOx [g/kWh]
uted to the evaporation of the ethanol, generally called cooling
effect [32,56]. The fuelling and injection strategy also resulted in Fig. 7. NOx/soot emissions trade-off.
a single stage heat release, differing from the typical double-
hump of conventional diesel combustion. Finally, the faster and
more advanced combustion process of the dual-fuel mode led to of the burnt gas temperature would require appropriate modelling
higher peak pressure (Pmax) and higher average combustion of the fuel distribution and heat release of the premixed and
temperatures than the diesel-only operation. The determination non-premixed regions which is out of the scope of this work.
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182 173

Fig. 8. Combustion and net indicated efficiencies.

Combustion phasing could still be controlled by the injection

timing without the need for large amounts of EGR.
Table 4 compares the resulting performance and emissions of
the two combustion modes. In dual-fuel mode, the longer ignition
delay and the reduced amount of high reactivity fuel minimised
fuel-rich and high temperatures zones [26,28]. As a result, NOx
and soot emissions were reduced by more than 50% when com-
pared to the conventional diesel combustion case. The faster and
more advanced combustion process also lowered heat transfer
losses [28] and allowed for higher indicated efficiency. Combustion
efficiency dropped as a result of reduced fuel reactivity and more Fig. 9. In-cylinder pressure, diesel injection, HRR, and average in-cylinder gas
premixed fuel trapped in the crevice and squish volumes [53]. As temperature curves. Optimum DF combustion strategy vs. equivalent diesel-only
the Euro VI emissions standards for NOx have not been fulfilled, case.

investigations with EGR were performed in the next section.

Table 4
4.3. The effect of EGR in DF combustion Comparison between the optimum DF combustion strategy and an equivalent diesel-
only case running with an intake air pressure of 103 kPa, an RP of 70 MPa, and no EGR.
This section presents the results of the experiments conducted Parameter Unit Diesel DF
to evaluate the effect of EGR in the optimum DF combustion
EF – 0.00 0.54
strategy. The boundary conditions were kept constant with the SOI_1 CAD ATDC 14.5 36.1
exception of intake manifold air temperature, which was elevated ET_1 ms 0.45 0.90
to 306 ± 1 K due to the addition of 25% EGR at 343 ± 5 K. The etha- SOI_2 CAD ATDC 7.3 0.2
nol mass flow rate was kept at 53.6 mg/inj. As the control of com- Split ratio % 25/75 60/40
Ignition delay – SOI_1 to SOC ms 1.67 3.20
bustion phasing in the DF mode relies on the diesel injection COV_IMEP % 1.2 1.4
strategy, SOI_1 was fixed at around 36.5 CAD ATDC with a con- Pmax MPa 7.12 8.67
stant energising timing ET_1 of 0.90 ms, providing a split ratio of PRR MPa/CAD 1.07 0.94
approximately 60/40. Experiments were carried out first with con- CA50 CAD ATDC 7.0 2.1
CA10–CA90 CAD 28.5 13.5
stant SOI_2 and then at similar CA50, by advancing SOI_2. Fig. 10
Uglobal – 0.48 0.44
shows the in-cylinder pressure, diesel injection, and HRR curves ISSoot g/kW h 0.031 0.011
of the optimum DF strategy at 0% and 25% EGR. Table 5 ISNOx g/kW h 7.85 3.56
summarises the resulting performance and emissions of the two ISCO g/kW h 0.80 4.56
experiments with and without EGR. ISHC g/kW h 0.42 4.90
gc % 99.7 97.4
In the case of constant SOI_2, adding EGR delayed combustion g_ind % 43.6 45.5
into the expansion stroke and increased the combustion duration.
174 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

peak heat release. Net indicated efficiency with EGR was slightly
lower than that without EGR, possibly as a consequence of non-
optimised combustion phasing and higher global equivalence ratio.
As the Euro VI NOx and soot emissions targets have not been
reached, further experiments were carried out by varying the first
diesel injection timing and split ratio at a constant SOI_2, as
described in the following section.

4.4. Optimisation of the SOI_1 and split ratio at a constant SOI_2

Sensitivity studies of the first diesel injection timing and split

Fig. 10. In-cylinder pressure, diesel injection, and HRR curves of the optimum DF
strategy at 0% and 25% EGR.
CA50 was retarded from 2.1 to 10.6 CAD ATDC and CA10–CA90 was
extended from 13.5 to 17 CAD, drastically reducing NOx
emissions from 3.56 to 0.69 g/kW h. Soot, CO, and unburnt HC
emissions increased to a certain extent due to late and poor
combustion development. Diesel fuel consumption also increased
with a more diluted charge in order to keep the engine speed
and IMEP constants, reducing the ethanol energy fraction from
54% to 51%.
In the second case, the SOI_2 was advanced to phase the MFB
profile closer to the CA50 of the 0% EGR case. The introduction of
EGR allowed for a longer ignition delay. Once the combustion
started, a more readily ignitable charge burnt in almost half the
time. NOx emissions were curbed by EGR. CO and unburnt HC
emissions decreased as a result of longer mixing period and higher
ratio on combustion, emissions, and efficiency of the optimum
DF strategy were carried out in this section. This is prior to testing
a higher rail pressure and a lower equivalence ratio, which are
addressed in the subsequent section. The operating conditions for
this series of experiments were the same as Section 4.3 at an
EGR rate of 25%. Unlike the constant dwell timing experiments per-
formed in Section 4.2, SOI_1 and ET_1 were varied while the SOI_2
was held at approximately 7.5 CAD ATDC. This provided a safety
margin to PRR and CA50. The injection pressure was set at 70 MPa.
The ethanol mass flow rate was kept at 53.6 mg/inj, resulting in an
energy fraction varying from 50% to 53%.
Fig. 11 depicts the SOI_1 sweeps at a split ratio of approxi-
mately 57/43. A maximum PRR of 1.71 MPa/CAD was found with
an SOI_1 occurring at 40 CAD ATDC. It was also the calibration
with the shortest combustion duration of 7.6 CAD. Earlier pre-
injections reduced fuel-rich zones, which was supported by a drop
in soot emissions and increase in COV_IMEP. The result was a
slower and retarded combustion process, increasing combustion
losses and hindering the indicated efficiency. Pre-injections later
than 40 CAD ATDC reduced the time available for mixing and
regions of higher local equivalence ratio prevailed. This stratifica-
tion advanced the CA50 and increased the peak in-cylinder pres-
sure and temperatures, resulting in higher NOx formation.
Another consequence of over advanced combustion phasing was
the increase in the compression work followed by a reduction in
net indicated efficiency. Accordingly, the optimum emissions/effi-
ciency trade-off was obtained by an SOI_1 taking place around
36.5 CAD ATDC.
Fig. 12 shows the results of the split ratio sweep performed
with a constant SOI_1 at 36.5 CAD ATDC. The ET_1 was varied
from 0.80 to 1.01 ms, representing 49–64% of the total energising
time. Increasing the fuel quantity injected in the first injection
yielded similar effects to retarding SOI_1. At the largest first injec-
tion amount (i.e. split ratio of 64/36), the more reactive mixture led

Table 5
The effect of EGR on combustion, emissions, and efficiency of the optimum DF strategy running at an intake air pressure of 103 kPa and an RP of 70 MPa.

Parameter Unit 0% EGR 25% EGR, similar SOI_1 and SOI_2 25% EGR, similar SOI_1 and CA50
EF – 0.54 0.51 0.54
SOI_1 CAD ATDC 36.1 36.6 36.6
ET_1 ms 0.90 0.90 0.90
SOI_2 CAD ATDC 0.2 0.2 9.8
Split ratio % 60/40 57/43 57/43
Ignition delay – SOI_2 to SOC ms 1.83 0.75 0.29
COV_IMEP % 1.4 1.5 1.1
Pmax MPa 8.67 6.45 9.13
PRR MPa/CAD 0.94 0.56 1.82
CA50 CAD ATDC 2.1 10.6 1.9
CA10–CA90 CAD 13.5 17.0 6.9
Uglobal – 0.44 0.63 0.60
ISSoot g/kW h 0.011 0.018 0.013
ISNOx g/kW h 3.56 0.69 2.73
ISCO g/kW h 4.56 5.86 3.26
ISHC g/kW h 4.90 5.41 3.62
gc % 97.4 97.1 98.1
g_ind % 45.5 44.2 45.1
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182 175

0.46 8

ISNOx, ISCO, ISHC [g/kWh]

6
0.44
_ind [-]

0.42
Net indicated eff.
2
CA10-CA90
CA50 ISNOx
PRR ISCO
0.40 18 0 ISHC 0.04
ISSoot
COV

CA50 [CAD ATDC]

COV_IMEP [%/100]
CA10-CA90 [CAD]
0.03

PRR [MPa/CAD]

ISSoot [g/kWh]
10

0.02

2
0.01

-6 0.00
-50 -45 -40 -35 -30 -25 -50 -45 -40 -35 -30 -25
SOI_1 [CAD ATDC] SOI_1 [CAD ATDC]

Fig. 11. Effect of the SOI_1 on combustion, emissions, and efficiency.

0.46 7.5
ISNOx, ISCO, ISHC [g/kWh]

0.44 5.0
_ind [-]

Net indicated eff.

CA10-CA90
0.42 2.5
CA50
PRR
ISNOx
ISCO
0.40 12 0.0 0.03
ISHC
ISSoot
COV
COV_IMEP [%/100]
CA50 [CAD ATDC]
CA10-CA90 [CAD]

PRR [MPa/CAD]

ISSoot [g/kWh]

6 0.02

0 0.01

-6 0.00
45% 50% 55% 60% 65% 70% 45% 50% 55% 60% 65% 70%
ET_1/total energising time [%] ET_1/total energising time [%]

Fig. 12. Effect of split ratio on combustion, emissions and efficiency.

to a rapid and early combustion, reducing unburnt HC and CO
emissions at the expense of higher NOx and PRR. Soot also
increased as the mixing time available to the main injection was
reduced. The opposite was true for decreasing the ET_1, however
a trade-off in indicated efficiency was observed. A short
pre-injection (i.e. split ratio of 49/51) reduced the mixture flamma-
bility and delayed the combustion process, degrading fuel conver-
sion efficiency. The optimum split ratio represents an ET_1
equivalent to 53–55% of the total energising time. The COV_IMEP
remained below 1.4%.
176 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

4.5. The effect of higher intake air pressure and rail pressure in DF
combustion

In this section, dual-fuel experiments were carried with higher

intake air pressure and diesel injection pressure. The objective was
to improve combustion and indicated efficiencies while reducing
soot and NOx emissions. Plots of two different rail pressures (70
and 90 MPa) and intake manifold air pressures (103 and 125 kPa)
were compared on a CA50 basis. The ethanol energy fraction varied
from 51% to 54%, with an average of 53%, as the amounts of diesel
fuel were automatically adjusted by the ECU to maintain a con-
stant engine speed and load. The SOI_1 was set at around 36.5
CAD ATDC with an ET_1 equivalent to a split ratio of approximately
54/46. The SOI_2 was altered within the range 9.5 to 1.0 CAD
ATDC to delay or advance the CA50. The EGR was introduced into
the system at 353 ± 10 K and maintained at 25.0 ± 0.5%. The result-
ing intake air charge temperature was 308 ± 2 K. The pumping
losses were the same, as the delta between the intake air pressure
and the exhaust back pressure was held at 10 kPa. Combustion sta-
bility was considered acceptable, with COV_IMEP between 0.9%
and 2.2%.
Fig. 13 depicts the in-cylinder pressure, diesel injection, and
HRR curves for a sweep of SOI_2 at 125 kPa intake air pressure
and 90 MPa injection pressure. Although the SOC’s took place
almost at the same crank angle, a retarded SOI_2 shifted the com-
bustion process towards the expansion stroke, decreasing the peak
in-cylinder pressure and peak HRR.
To expand on this trend, SOI_2, combustion duration, ignition
delay, and PRR with respect to CA50 are shown in Fig. 14. The dif-
fusion combustion of the second diesel injection allowed control
over the combustion phasing and burn rate. A higher intake air
pressure required a retarded SOI_2 due earlier auto-ignition of
the diesel fuel injected at the 36.5 CAD ATDC, suggested by the
shorter ignition delays. Earlier main injections generally allowed
for more time available between the SOI_2 and the SOC. However,

Fig. 14. SOI_2, ignition delay, CA10–CA90, and PRR vs. CA50.

Fig. 13. In-cylinder pressure, diesel injection, and HRR curves for a sweep of SOI_2.
the most advanced cases at 125 kPa intake air pressure presented
short ignition delays (i.e. highlighted by the circled region). The
higher in-cylinder pressures and temperatures advanced the onset
of auto-ignition of the premixed diesel (i.e. ET_1). Despite the
earlier SOC, an intake air pressure of 125 kPa possesses the longest
CA10–CA90 and the lowest PRR. This is a result of the dilution with
excess air, which diminishes the reaction rates [57]. An injection
pressure of 90 MPa resulted in better atomisation and more
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182 177

Fig. 15. Effect of intake air pressure and diesel injection pressure in DF combustion.

homogeneous charge prior to the SOI_2, increasing the PRR and

shortening the combustion process. The PRR limit of 2.0 MPa/
CAD was exceeded during the most advanced cases at 103 kPa
intake air pressure.
Fig. 15 shows the effect of higher boost and diesel injection
pressures on in-cylinder pressure and the resulting heat release
at a CA50 of approximately 5.8 CAD ATDC. A lower rail pressure
required a retarded SOI_2 to obtain the same combustion phasing
because of the formation of high equivalence ratio regions within
the piston bowl. Another important observation concerns the
HRR profile of the DF combustion at 103 kPa intake air pressure
and 90 MPa injection pressure. The combination of lower in-
cylinder pressure and improved atomisation resulted in longer
ignition delay and elevated PRR. It is believed that the two distinct
high temperature heat release events are a result of the fast burn-
ing of the second diesel injection, followed by the ignition of the
entrained premixed charge. Higher intake air pressure and lower
RP allowed for a more progressive HRR.
Fig. 16 depicts the effect of intake air pressure and injection
pressure on emissions and efficiencies of DF combustion. Despite
the higher in-cylinder pressures prior to the SOC at 125 kPa intake
air pressure, the dilution with excess air constrained the charge
reactivity and heat release peaks, mitigating NOx formation. The
opposite occurs if rail pressure is increased as a result of better
atomisation (i.e. presence of close-to-stoichiometric regions) and
Fig. 16. The effect of intake air pressure and injection pressure on emissions and
faster combustion process. The improved mixture preparation also
efficiencies of DF combustion.
helps with fuel efficiency gains and smoke reduction. When com-
paring the emissions on a CA50 basis, an advanced SOI_2 resulted
in higher NOx emissions due to shorter CA10–CA90 and elevated Combustion efficiency in DF mode remains lower than conven-
in-cylinder peak pressures and temperatures. In addition, an ear- tional diesel combustion (i.e. generally above 99%). It is attributed
lier SOI_2 at an intake air pressure of 103 kPa led to higher soot to the reduced reactivity of the ethanol and the fundamental differ-
emissions after poor mixing and rapid combustion. The operation ences between a turbulent diffusion flame in diesel-only operation
at 125 kPa exhibited the contrary behaviour, with soot emissions and a progressive and sequential combustion from the high to low
increasing as the SOI_2 was retarded. This is a consequence of a reactivity regions in DF mode [1]. The diesel consumption does not
poor combustion process and the formation of low temperature progress to the crevice regions, not releasing enough energy for the
fuel-rich regions. combustion of all the premixed fuel, which increases combustion
178 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

an injection pressure of 125 kPa and 90 MPa, respectively. Fig. 17

Fig. 17. In-cylinder pressure, diesel injection, and MFB curves of the optimum
emissions trade-off in diesel-only and DF combustion modes compared against the
most fuel efficient case.
losses [53]. The use of an intake air pressure of 103 kPa and earlier
SOI_2’s improved the flammability of the in-cylinder charge and
resulted in combustion efficiencies up to 98%, at the expense of
higher NOx emissions. A higher intake air pressure and a retarded
SOI_2 reduced NOx emissions and combustion efficiencies due to
overly lean regions and lower local in-cylinder temperatures.
4.6. Comparison between DF and conventional diesel combustion
The optimum dual-fuel operating points were determined by
the best indicated efficiency/emissions trade-off at an intake and
compares the in-cylinder pressure, injector signal, and MFB curves
of the optimum emissions trade-off in diesel-only (see Section 4.1)
and DF combustion against the most fuel efficient case attained
during this study. As showed in Section 4.2, dual-fuel mode is char-
acterised by advanced combustion phasing, shorter combustion
durations, and higher peak in-cylinder pressures. This is due to
the early first diesel injection which increases the flammability of
the in-cylinder charge and promotes a more reactive mixture prior
to the second injection. The main diesel injection close to TDC
determines the combustion phasing. It is believed that advanced
SOI_2 will be required at higher ethanol substitution ratios to
increase the degree of stratification and avoid misfire.
Table 6 compares the resulting performance and emissions of
the optimum conventional diesel combustion baseline (deter-
mined in Section 4.1) against the two selected dual-fuel operating
points. The first DF strategy represents the operating point with
optimum g_ind/ISNOx/ISSoot trade-off, while the second is the
most fuel efficient calibration obtained at this particular load. It
can be observed that a premixed charge of ethanol resulted in sev-
eral benefits. These included higher indicated efficiency, and lower
soot and NOx emissions than diesel-only operation. The use of
optimised split diesel injections, intake air pressure, and rail pres-
sure also kept ISCO and ISHC below 10 g/kW h, demonstrating a
considerable improvement in comparison to previous studies
[37–39]. The higher indicated efficiency associated with dual-fuel
combustion led to lower exhaust gas temperatures (EGT) than con-
ventional diesel combustion [58]. Elevated injection pressure of
125 MPa was required to curb soot emissions in diesel-only mode.
The radar chart showed in Fig. 18 summarises the main trends
observed. Net indicated efficiency, peak in-cylinder pressure, soot
and NOx emissions, and combustion efficiency results were
normalised to obtain a clear distinction of the benefits. Euro VI
legislation emissions limits were not fully met under conventional
diesel combustion or dual-fuel mode. However, the best results
were obtained with utilisation of ethanol. The combustion of a
well-mixed charge ignited by split diesel injections considerably
reduced NOx and soot emissions, which are the main limiting
factors of HD diesel engines. Net indicated efficiency increased
by 3.2% in the most efficient case as a result of lower local
in-cylinder temperatures and faster combustion process. The
majority of the unburnt HC and CO emissions produced by DF
operation can be removed by an oxidation catalyst, assuming an
EGT above 570 K [58].

Table 6
Comparison between the optimum diesel-only calibration and the selected dual-fuel operating points, running with 25% EGR and 125 kPa intake air pressure.

Parameter Unit Diesel, trade-off DF, trade-off DF, most efficient

EF – 0.00 0.53 0.54
SOI_1 CAD ATDC 9.3 36.7 36.7
ET_1 ms 0.38 0.76 0.76
SOI_2 CAD ATDC 2.0 0.8 2.7
Split ratio % 30/70 53/47 55/45
Rail pressure MPa 125 90 90
Ignition delay – SOI_2 to SOC ms 0.49 1.12 0.77
COV_IMEP % 1.8 1.6 1.4
Pmax MPa 7.56 7.85 9.09
PRR MPa/CAD 1.42 0.67 1.26
CA50 CAD ATDC 9.2 7.8 4.1
CA10–CA90 CAD 22.3 18.6 14.6
EGT K 596 578 567
Uglobal – 0.53 0.51 0.50
ISSoot g/kW h 0.0175 0.0125 0.0118
ISNOx g/kW h 2.01 0.71 1.32
ISCO g/kW h 0.67 9.55 6.94
ISHC g/kW h 0.14 6.69 5.34
gc % 99.8 96.0 96.8
g_ind % 44.1 44.4 45.5
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182
Pmax
+0.7%
_ind
+3.2%
-4%
-3%
c ISNOx
Fig. 18. Normalised net indicated efficiency, peak in-cylinder pressure, ISSoot and ISNOx emissions, and combustion efficiency results of the optimum trade-offs in diesel-
only and DF combustion modes compared against the most fuel efficient case.
5. Conclusions
In this paper, ethanol dual-fuel combustion was experimentally
investigated using an HD diesel engine operating at 1200 rpm and
0.615 MPa. The impact of three ethanol fractions and different die-
sel injection strategies on combustion, emissions, and efficiency
were analysed and discussed. Two separate engine calibrations
with the highest efficiency and lowest emissions were quantified,
complete with effects from EGR, intake air pressure, and injection
pressure. The main findings can be summarised as follows:

Diesel-only combustion requires a combination of very high
injection pressures and EGR rates to achieve low engine-out
emissions of soot and NOx. Intake air pressure also needs to
be increased to avoid a fuel economy penalty.

Ethanol dual-fuel combustion with a single diesel injection
close to TDC has a limited operating range due to high PRR
and low indicated efficiency. A split injection strategy adjusted
the mixture flammability and promoted in-cylinder reactivity
gradients, increasing the fuel conversion efficiency and reduc-
ing CO and unburnt HC emissions.

In the majority of the cases tested, the homogenous charge of
ethanol reduced local in-cylinder temperatures and fuel-rich
zones, resulting in lower NOx and soot emissions compared to
conventional diesel combustion.

High ethanol substitution ratios resulted in poor combustion
efficiency at this particular load due low in-cylinder charge
temperature and high amount of fuel trapped in the crevice
and squish volumes. Low substitution ratios did not demon-
strate benefits in terms of net indicated efficiency and emis-
sions reduction.

The optimum DF strategy was an ethanol substitution ratio of
54% on an energy basis, with a diesel pre-injection timing
around 36.5 CAD ATDC and a split ratio of corresponding to
53–55% of the total energising time.

The addition of EGR reduced NOx emissions and promoted
longer ignition delays, consequently allowing more time for
the charge to mix prior to the SOC.

The split diesel injection strategy used in this work allowed for
control over combustion phasing without the need for varying
the overall fuel reactivity (i.e. ethanol substitution ratio). The
179
Diesel, trade-off
+20%
EF 0.53, trade-off
+3.8%
EF 0.54, most fuel efficient
ISSoot
-33%
-29%
-65%
-34%
use of a second diesel injection around TDC rather than early
injections is a differentiator from RCCI.

Reducing rail pressure in dual-fuel mode created a higher
degree of stratification, advancing the SOC. The shorter mix-
ing period resulted in longer diffusion combustion and higher
soot emissions, but lower NOx levels. A higher injection
pressure improved diesel atomisation and resulted in a more
homogeneous charge, delaying the SOC. This contributed to
shorter combustion process, higher PRR, and higher NOx
emissions.

A leaner operation (higher intake air pressure) decreased local
in-cylinder temperatures and heat release peaks. This resulted
in lower NOx formation, but higher unburnt HC and CO
emissions.
Ethanol dual-fuel combustion offers the potential to minimise
exhaust aftertreatment requirements and improve the efficiency
of heavy-duty engines. Furthermore, the use of ethanol allows for
less dependence on petroleum-based fuels and a reduction in
greenhouse gas emissions [5]. While Euro VI emissions standards
still remain a challenge, the calibration for the best dual-fuel emis-
sions case reduced NOx emissions by 65% and soot emissions by
29% when compared against the optimum conventional diesel
combustion mode. The most efficient dual-fuel calibration allowed
for a 3.2% gain in indicated efficiency while retaining at least 33%
reduction in NOx and soot emissions. Combustion losses were mit-
igated by maintaining CO and unburnt HC under 10 g/kW h, but a
low temperature catalyst may be required [25]. Dual-fuel opera-
tion at higher loads might be limited by the maximum cylinder
pressure of the engine. Future work will be carried out at different
engine speeds and loads to maximise the substitution of ethanol
while minimising emissions.
Acknowledgments
The authors would like to acknowledge the Brazilian Federal
Agency for Support and Evaluation of Postgraduate Education
(CAPES) and the National Council for Scientific and Technological
Development (CNPq) for supporting the PhD studies of Mr. Pedrozo
and Mr. Dalla Nora at Brunel University London.
180 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

Appendix A. Measurement device specifications

Measured variable Device Manufacturer Dynamic Linearity/ Repeatability

range accuracy
CO (low content) AIA-721A Horiba (MEXA 7170 0–2.5k ppm 6 ±1.0% FS or Within ±0.5% of
CO (mid-high content) AIA-722 DEGR) 0–12 vol% ±2.0% of FS
CO2 AIA-722 0–20 vol% readings
NOx CLA-720MA 0–500 ppm or
0–10k ppm
O2 MPA-720 0–25 vol%
Unburnt HC FIA-725A 0–500 ppm or
0–50k ppm
Diesel injector current signal Current Probe PR30 LEM 0–20 A ±1% of reading
±2 mA
Diesel flow rate (return) PROline promass 83A Endress + Hauser 0–100 kg/h ±0.10% of ±0.05% of
DN01 reading reading
Diesel flow rate (supply) PROline promass 83A 0–20 kg/h ±0.10% of ±0.05% of
DN02 reading reading
Intake and exhaust pressures Piezoresistive Kistler 0–1 MPa 6 ±0.50% of FS
pressure sensor Type within 0–353 K
4049A
Amplifier Type
4622A
In-cylinder pressure Piezoelectric Kistler 0–30 MPa 6 ±0.40% of FS
pressure sensor Type
6125C
Amplifier FI Piezo AVL 6 ±0.01% of FS
Intake valve lift S-DVRT-24 LORD MicroStrain 0–24 mm ±1% of reading ±1.0 lm
displacement sensor using straight
DEMOD-DVRT-TC line
conditioner
Intake air mass flow rate Proline t-mass 65F Endress + Hauser 0–910 kg/h ±1.5% of reading ±0.5% of reading
(10–100% of FS)
Oil and ethanol pressure Pressure transducer GE 0–1 MPa < ±0.20% of FS
UNIK 5000
Smoke number 415SE AVL 0–10 FSN – Within ±0.005
FSN +3% of
reading
Speed AG150 Froude Hofmann 0–8000 rpm ±1 rpm
Torque Dynamometer 0–500 Nm ±0.25% of FS
Temperature Thermocouple K RS 233–1473 K 6± 2.5 K or
Type (Class 2) ±0.75% of
readings
⁄ FS = full scale.

References
[1] Reitz RD. Directions in internal combustion engine research. Combust Flame
2013;160:1–8. http://dx.doi.org/10.1016/j.combustflame.2012.11.002.
[2] Dec JE. A conceptual model of DI diesel combustion based on laser sheet
imaging. SAE Technical Paper 1997. http://dx.doi.org/10.4271/970873.
[3] Imtenan S, Varman M, Masjuki HH, Kalam Ma, Sajjad H, Arbab MI, et al. Impact
of low temperature combustion attaining strategies on diesel engine emissions
for diesel and biodiesels: a review. Energy Convers Manage 2014;80. http://dx.
doi.org/10.1016/j.enconman.2014.01.020.
[4] U.S. Energy Information Adminstration. International Energy Outlook 2014.
Washington; 2014.
[5] United Nations Environment Programme. Towards sustainable production and
use of resources: assessing biofuels; 2009.
[6] Asad U, Zheng M, Ting DS-K, Tjong J. Implementation challenges and solutions
for homogeneous charge compression ignition combustion in diesel engines. J
Eng Gas Turbines Power 2015;137. http://dx.doi.org/10.1115/1.4030091.
[7] Bendu H, Murugan S. Homogeneous charge compression ignition (HCCI)
combustion: mixture preparation and control strategies in diesel engines.
Renew Sustain Energy Rev 2014;38. http://dx.doi.org/10.1016/j.
rser.2014.07.019.
[8] Yao M, Zheng Z, Liu H. Progress and recent trends in homogeneous charge
compression ignition (HCCI) engines. Prog Energy Combust Sci
2009;35:398–437. http://dx.doi.org/10.1016/j.pecs.2009.05.001.
[9] Kook S, Bae C. Combustion control using two-stage diesel fuel injection in a
single-cylinder PCCI engine. SAE Technical Paper 2004. http://dx.doi.org/10.
4271/2004-01-0938.
[10] Neely GD, Sasaki S, Huang Y, Leet JA, Stewart DW. New diesel emission control
strategy to meet US tier 2 emissions regulations. SAE Technical Paper 2005.
http://dx.doi.org/10.4271/2005-01-1091.
[11] Kook S, Bae C, Miles PC, Choi D, Pickett LM. The influence of charge dilution
and injection timing on low-temperature diesel combustion and emissions.
SAE Technical Paper 2005. http://dx.doi.org/10.4271/2005-01-3837.
[12] Opat R, Ra Y, Gonzalez D. MA, Krieger R, Reitz RD, Foster DE, et al. Investigation
of mixing and temperature effects on HC/CO emissions for highly dilute low
temperature combustion in a light duty diesel engine. SAE Technical Paper
2007. http://dx.doi.org/10.4271/2007-01-0193.
 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182
[13] Musculus MPB, Miles PC, Pickett LM. Conceptual models for partially premixed
low-temperature diesel combustion. Prog Energy Combust Sci 2013;39. http://
dx.doi.org/10.1016/j.pecs.2012.09.001.
[14] Kimura S, Aoki O, Ogawa H, Muranaka S, Enomoto Y. New combustion concept
for ultra-clean and high-efficiency small DI diesel engines. SAE Technical Paper
1999. http://dx.doi.org/10.4271/1999-01-3681.
[15] Hasegawa R, Yanagihara H. HCCI combustion in DI diesel engine. SAE Technical
Paper 2003. http://dx.doi.org/10.4271/2003-01-0745.
[16] Sellnau M, Foster M, Hoyer K, Moore W, Sinnamon J, Husted H. Development of
a gasoline direct injection compression ignition (GDCI) engine. SAE Int J Eng
2014;7:835–51. http://dx.doi.org/10.4271/2014-01-1300.
[17] Sellnau M, Moore W, Sinnamon J, Hoyer K, Foster M, Husted H. GDCI multi-
cylinder engine for high fuel efficiency and low emissions. SAE Int J Eng
2015;8. http://dx.doi.org/10.4271/2015-01-0834.
[18] Kalghatgi GT, Risberg P, Ångström H-E. Partially pre-mixed auto-ignition of
gasoline to attain low smoke and low NOx at high load in a compression
ignition engine and comparison with a diesel fuel. SAE Technical Paper 2007.
http://dx.doi.org/10.4271/2007-01-0006.
[19] Manente V, Zander C, Johansson B, Tunestal P, Cannella W. An advanced
internal combustion engine concept for low emissions and high efficiency
from idle to max load using gasoline partially premixed combustion. SAE
Technical Paper 2010. http://dx.doi.org/10.4271/2010-01-2198.
[20] Manente V, Johansson B, Cannella W. Gasoline partially premixed combustion,
the future of internal combustion engines? Int J Engine Res 2011;12:194–208.
http://dx.doi.org/10.1177/1468087411402441.
[21] Manente V, Johansson B, Tunestal P. Characterization of partially premixed
combustion with ethanol: EGR sweeps, low and maximum loads. J Eng Gas
Turbines Power 2010;132. http://dx.doi.org/10.1115/1.4000291.
[22] Shen M, Tuner M, Johansson B, Cannella W. Effects of EGR and intake pressure
on PPC of conventional diesel, gasoline and ethanol in a heavy duty diesel
engine. SAE Technical Paper 2013;01. http://dx.doi.org/10.4271/2013-01-
2702.
[23] Shen M, Tuner M, Johansson B. Close to stoichiometric partially premixed
combustion – the benefit of ethanol in comparison to conventional fuels. SAE
Technical Paper 2013. http://dx.doi.org/10.4271/2013-01-0277.
[24] Kaiadi M, Johansson B, Lundgren M, Gaynor JA. Experimental investigation on
different injection strategies for ethanol partially premixed combustion. SAE
Technical Paper 2013. http://dx.doi.org/10.4271/2013-01-0281.
[25] Reitz RD, Duraisamy G. Review of high efficiency and clean reactivity
controlled compression ignition (RCCI) combustion in internal combustion
engines. Prog Energy Combust Sci 2015;46:12–71. http://dx.doi.org/10.1016/j.
pecs.2014.05.003.
[26] Benajes J, Molina S, García A, Belarte E, Vanvolsem M. An investigation on RCCI
combustion in a heavy duty diesel engine using in-cylinder blending of diesel
and gasoline fuels. Appl Therm Eng 2014;63:66–76. http://dx.doi.org/10.1016/
j.applthermaleng.2013.10.052.
[27] Kokjohn SL, Hanson R, Splitter D, Kaddatz J, Reitz RD. Fuel reactivity controlled
compression ignition (RCCI) combustion in light- and heavy-duty engines. SAE
Int J Fuels Lubr 2011;4. http://dx.doi.org/10.4271/2011-01-0357.
[28] Kokjohn SL, Hanson RM, Splitter Da, Reitz RD. Fuel reactivity controlled
compression ignition (RCCI): a pathway to controlled high-efficiency clean
combustion. Int J Engine Res 2011;12:209–26. http://dx.doi.org/10.1177/
1468087411401548.
[29] Sarathy SM, Oßwald P, Hansen N, Kohse-Höinghaus K. Alcohol combustion
chemistry. Prog Energy Combust Sci 2014;44:40–102. http://dx.doi.org/
10.1016/j.pecs.2014.04.003.
[30] Martins MES, Lanzanova TDM. Full-load Miller cycle with ethanol and EGR:
potential benefits and challenges. Appl Therm Eng 2015;90:274–85. http://dx.
doi.org/10.1016/j.applthermaleng.2015.06.086.
[31] He B, Wang J, Shuai S, Yan X. Homogeneous charge combustion and emissions
of ethanol ignited by pilot diesel on diesel engines. SAE Technical Paper 2004.
http://dx.doi.org/10.4271/2004-01-0094.
[32] Gao T, Divekar P, Asad U, Han X, Reader GT, Wang M, et al. An enabling study of
low temperature combustion with ethanol in a diesel engine. J Energy Res
Technol 2013;135. http://dx.doi.org/10.1115/1.4024027.
[33] Han X, Xie K, Tjong J, Zheng M. Empirical study of simultaneously low NOx and
soot combustion with diesel and ethanol fuels in diesel engine. J Eng Gas
Turbines Power 2012;134. http://dx.doi.org/10.1115/1.4007163.
[34] Divekar P, Yang Z, Ting D, Chen X, Zheng M, Tjong J. Efficiency and emission
trade-off in diesel-ethanol low temperature combustion cycles. SAE Technical
Paper 2015. http://dx.doi.org/10.4271/2015-01-0845.
[35] Han X, Zheng M, Tjong J. Clean combustion enabling with ethanol on a dual-
fuel compression ignition engine. Int J Engine Res 2015:1–13. http://dx.doi.
org/10.1177/1468087415575646.
[36] Han X, Divekar P, Reader G, Zheng M, Tjong J. Active injection control for
enabling clean combustion in ethanol-diesel dual-fuel mode. SAE Int J Engines
2015;8:890–902. http://dx.doi.org/10.4271/2015-01-0858.
[37] Asad U, Kumar R, Zheng M, Tjong J. Ethanol-fueled low temperature
combustion: a pathway to clean and efficient diesel engine cycles. Appl
Energy 2015. http://dx.doi.org/10.1016/j.apenergy.2015.01.057.
[38] Júnior RFB, Martins CA. Emissions analysis of a diesel engine operating in
Diesel–Ethanol Dual-Fuel mode. Fuel 2014;148:191–201. http://dx.doi.org/
10.1016/j.fuel.2014.05.010.
[39] Ogawa H, Shibata G, Kato T, Zhao P. Dual fuel diesel combustion with
premixed ethanol as the main fuel. SAE Technical Paper 2014. http://dx.doi.
org/10.4271/2014-01-2687.
181
[40] Sarjovaara T, Alantie J, Larmi M. Ethanol dual-fuel combustion concept on
heavy duty engine. Energy 2013;63:76–85. http://dx.doi.org/10.1016/j.
energy.2013.10.053.
[41] Padala S, Woo C, Kook S, Hawkes ER. Ethanol utilisation in a diesel engine
using dual-fuelling technology. Fuel 2013;109:597–607. http://dx.doi.org/
10.1016/j.fuel.2013.03.049.
[42] Heywood JB. Internal combustion engine fundamentals. 1st ed. McGraw-Hill
Inc.; 1988.
[43] Silvis WM. An algorithm for calculating the air/fuel ratio from exhaust
emissions. SAE Technical Paper 1997. http://dx.doi.org/10.4271/970514.
[44] Zhao H, Ladammatos N. Engine combustion instrumentation and diagnostics.
Warrendale (PA, USA); 2001.
[45] Kar K, Cheng WK. Speciated engine-out organic gas emissions from a PFI-SI
engine operating on ethanol/gasoline mixtures. SAE Int J Fuels Lubr 2009;2.
http://dx.doi.org/10.4271/2009-01-2673.
[46] Wallner T. Correlation between speciated hydrocarbon emissions and flame
ionization detector response for gasoline/alcohol blends. J Eng Gas Turbines
Power 2011;133. http://dx.doi.org/10.1115/1.4002893. 082801–082801 – 8.
[47] AVL. AVL 415SE smoke meter – product guide. Graz, Austria; 2013.
[48] Economic Commission for Europe of the United Nations (UN/ECE). Regulation
No 49 – uniform provisions concerning the measures to be taken against the
emission of gaseous and particulate pollutants from compression-ignition
engines and positive ignition engines for use in vehicles. Off J Eur Union 2013.
[49] Splitter DA, Reitz RD. Fuel reactivity effects on the efficiency and operational
window of dual-fuel compression ignition engines. Fuel 2014;118. http://dx.
doi.org/10.1016/j.fuel.2013.10.045.
[50] Zhao H. HCCI and CAI engines for automotive industry. Cambridge: Woodhead
Publishing Limited; 2007.
[51] Tutak W, Lukács K, Szwaja S, Bereczky Á. Alcohol–diesel fuel combustion in the
compression ignition engine. Fuel 2015;154:196–206. http://dx.doi.org/
10.1016/j.fuel.2015.03.071.
[52] May I, Cairns A, Zhao H, Pedrozo V, Wong HC, Whelan S, et al. Reduction of
methane slip using premixed micro pilot combustion in a heavy-duty natural
gas-diesel engine. SAE Technical Paper 2015. http://dx.doi.org/10.4271/2015-
01-1798.
[53] Desantes JM, Benajes J, García A, Monsalve-Serrano J. The role of the in-
cylinder gas temperature and oxygen concentration over low load reactivity
controlled compression ignition combustion efficiency. Energy
2014;78:854–68. http://dx.doi.org/10.1016/j.energy.2014.10.080.
[54] Kokjohn SL, Hanson RM, Splitter DA, Reitz RD. Experiments and modeling of
dual-fuel HCCI and PCCI combustion using in-cylinder fuel blending. SAE Int J
Engines 2009;2. http://dx.doi.org/10.4271/2009-01-2647.
[55] Sarjovaara T, Larmi M, Vuorinen V. Effect of charge air temperature on E85
dual-fuel diesel combustion. Fuel 2015;153:6–12. http://dx.doi.org/10.1016/
j.fuel.2015.02.096.
[56] Tutak W. Bioethanol E85 as a fuel for dual fuel diesel engine. Energy Convers
Manage 2014;86:39–48. http://dx.doi.org/10.1016/j.enconman.2014.05.016.
[57] Eichmeier J, Wagner U, Spicher U. Controlling gasoline low temperature
combustion by diesel micro pilot injection. J Eng Gas Turbines Power
2012;134. http://dx.doi.org/10.1115/1.4005997.
[58] Prikhodko VY, Curran SJ, Parks JE, Wagner RM. Effectiveness of diesel oxidation
catalyst in reducing HC and CO emissions from reactivity controlled
compression ignition. SAE Int J Fuels Lubr 2013;6. http://dx.doi.org/10.4271/
2013-01-0515. 2013–01 – 0515.
Glossary
ATDC: after firing top dead centre
CA10–CA90: combustion duration (10–90% cumulative heat release)
CA50: crank angle of 50% cumulative heat release
CAD: crank angle degree
CO: carbon monoxide
CO2: carbon dioxide
COV_IMEP: coefficient of variation of the IMEP
DAQ: data acquisition
DF: dual-fuel
ECR: effective compression ratio
ECU: engine control unit
EGR: exhaust gas recirculation
EGT: exhaust gas temperature
ESC13: European Stationary Cycle
ET: energising time
ET_1: energising time of the first diesel injection
ET_2: energising time of the second diesel injection
FID: flame ionisation detector
FSN: filter smoke number
GDCI: Gasoline Direct Injection Compression Ignition
GHG: greenhouse gas
HC: hydrocarbons
HCCI: Homogeneous Charge Compression Ignition
HD: heavy-duty
HPP: high-pressure diesel pump
HRR: apparent net heat release rate
182 V.B. Pedrozo et al. / Applied Energy 165 (2016) 166–182

ID: ignition delay PFI: port fuel injector

IMEP: net indicated mean effective pressure PPC: partially premixed combustion
ISFC: net indicated specific fuel consumption PPCI: partially premixed charge compression ignition
ISCO: net indicated specific emissions of CO RCCI: Reactivity Controlled Compression Ignition
ISHC: net indicated specific emissions of unburnt HC RP: rail pressure
ISNOx: net indicated specific emissions of NOx SOC: start of combustion
ISSoot: net indicated specific emissions of soot SOI: start of injection
IVC: intake valve closing SOI_1: start of the first diesel injection
LTC: low temperature combustion SOI_2: start of the second diesel injection
MFB: mass fraction burnt TDC: firing top dead centre
MK: Modulated Kinetics UNIBUS: Uniform Bulky Combustion System
NOx: nitrogen oxides VVA: variable valve actuation
O2: oxygen WHSC: World Harmonized Stationary Cycle
Pmax: peak in-cylinder pressure c: ratio of specific heats
PCCI: Premixed Charge Compression Ignition Uglobal: Global Equivalence Ratio