Comparative assessment of performance, emissions and combustion characteristics of gasoline/diesel and gasoline/biodiesel in a dual-fuel engine

Abstract

In recent years, rapid growth in population, development, and industrialization have led to a high demand for energy worldwide. Biofuels from bio-based products can be considered an alternative to fossil fuels used in the transport sector. However, the use of biodiesel in conventional diesel combustion engines has usually caused lower thermal efficiency and higher specific fuel consumption. Using alternative fuels and switching to promising combustion technologies such as low temperature combustion (LTC) are reliable approaches to address this issue. This research aims to use biofuels as an alternative energy source for engines operating in dual-fuel combustion mode. The effects of diesel/biodiesel strategies on dual-fuel combustion were investigated. This dual-fuel combustion mode proposes port fuel injection of gasoline and direct injection of diesel/biodiesel fuel with rapid in-cylinder fuel blending. The results of engine performance, emissions, and cylinder pressure were recorded and analyzed. The results showed that engines operating under dual-fuel combustion mode could achieve high efficiency with near zero nitrogen oxide (NO_x) and smoke emissions. The biodiesel–gasoline dual-fuel combustion mode showed lower hydrocarbon (HC) and carbon monoxide (CO) emissions than the diesel–gasoline dual-fuel combustion mode. The oxygen content in biodiesel is especially useful in limiting locally fuel rich regions, resulting in improved combustion and thereby reducing HC and CO emissions simultaneously.

---

1. Introduction

In the present scenario of a worldwide energy crisis coupled with its detrimental impact on environment, the world is compelled to focus on developing clean alternative fuel that is economically competitive, technically feasible, easily available, and environmentally acceptable.^1–3 Biofuels from bio-based products are an alternative to fossil fuels used in the transport sector. They are renewable, can improve the energy security of a country by reducing dependency on volatile foreign markets, and producing them can ease unemployment. As an alternative to diesel fuel, biodiesel is one of the most promising and ideal choices due to its environmentally adaptable behavior and similar physicochemical properties to those of fossil diesel fuel. Furthermore, its content is minimally toxic, highly biodegradable, has a higher cetane number, and an absence of aromatic compounds and sulfur, thus making it a more desirable alternative to diesel.⁴ Biodiesel can be derived from straight vegetable oils, edible and non-edible plants, recycled waste cooking oils, and animal fat.^5,6 Biodiesel in its neat or blended form can be used in diesel engines without modification of the engine or fueling process, thus greatly simplifying the system's integration and adoption.

Recently, another type of promising combustion strategies have evolved that are called low temperature combustion (LTC) in addition to spark ignition (SI) and CI. LTC is an in-cylinder approach of advanced combustion strategies for the simultaneous reduction of PM and NO_x emissions.^7,8 It can be adopted in any size of transportation engine, ranging from small,^9,10 light-duty^11,12 and heavy-duty engines,^13–15 and large ship engines.¹⁶ Various LTC concepts including homogenous charge compression ignition (HCCI),^17–23 uniform bulky combustion system (UNIBUS),²⁴ modulated kinetics (MK),²⁵ premixed charge compression ignition (PCCI),^26–28 homogeneous charge diesel combustion (HCDC),^29–31 homogeneous charge late injection (HCLI),³² and reactivity controlled compression ignition (RCCI),^33,34 are being intensively investigated as potential future alternatives for efficient internal combustion engines. Although various acronyms have been assigned to this new combustion process, they still refer to the common fundamental characteristics of a premixed fuel–air mixture and auto-ignited combustion, with the goal of lowering combustion temperatures to advantageously alter the chemistry of NO_x and/or soot formation.

The currently available LTC technologies, however, remains a formidable challenge due to their high HC, CO emissions, and narrow operating regimes that may related to complications in ignition control. Occasionally, the problems are serious enough to cause higher BSFC.^35,36 Generally, the low load limit is constrained by ignition stability and high cycle-to-cycle combustion instability, whereas high load operation is limited by excessively high maximum combustion pressure, rapid pressure rise and knocking combustion.³⁷ Therefore, a new emerging dual fuel engine combustion strategy, called RCCI by Kokjohn et al.³⁸ is worth investigating. RCCI is a dual fuel engine combustion technology and offers better control of combustion and resolve the load range limitation issue of HCCI and PCCI strategies.³⁹ As the name implies, the RCCI combustion approach employs in-cylinder fuel blending with at least two fuels of different reactivity injected at specific times in the engine cycle to control the in-cylinder charge reactivity and thus optimize combustion timing, duration, and magnitude. Usually, RCCI uses port injection of relatively low reactive fuel (i.e., gasoline) along with direct injection of higher reactive fuel (i.e., diesel) to control in-cylinder charge conditions. Mixing fuels of varied reactivity in the cylinder offers another powerful dimension of combustion control parameters. Reitz's investigations have demonstrated that an engine operating with RCCI strategy can gain back about 20% in fuel efficiency compared to conventional diesel combustion while still meeting PM and NO_x emissions without after-treatment.⁴⁰

1.1. Purpose of study

The use of biodiesel in conventional diesel combustion engines has usually caused higher NO_x and specific fuel consumption.^41–43 Using alternative fuels and switching to promising combustion technologies such as LTC can be reliable approaches to addressing this issue.⁴⁴ In fact, LTC is a promising concept for NO_x emission reduction not only for petroleum diesel, but also for biodiesels.⁴⁵ However, the main challenge for most LTC strategies is the higher HC and CO emissions that result from low combustion temperature and higher EGR rate. Using oxygenated fuel such as bioethanol and biodiesel can be a good alternative to this problem, and yet these fuels are derived from renewable sources. Furthermore, biofuels has received renewed interest due to its less polluting and renewable nature as opposed to petroleum fuels. However, very few studies have investigating engines operating on biofuels in RCCI combustion mode. In a RCCI dual-fuel combustion engine system, biodiesel fuel has a higher cetane number and higher oxygen content and can be used as the ignition source. Apparently, combining the two, LTC and biofuels, could potentially address both the emissions and efficiency challenges observed with petroleum–diesel based low temperature combustion.⁴⁶ Therefore, this research study focuses on the utilisation of biofuels as an alternative energy source for engines operating in RCCI dual-fuel combustion mode.

2. Experimental apparatus and procedure

2.1. Test fuels and operating conditions

In this study, fossil diesel fuel and palm oil were obtained in commercial form. Biodiesel selection in the present investigation was primarily based on ready availability of feedstock in Malaysia and their favorable fuel properties. Biodiesel production was conducted via the acid-esterification and alkali-transesterification process. Table 1 describes the key physicochemical of the converted neat palm methyl ester (PME) compared with ASTM and EN standards. The important properties of petroleum diesel and gasoline are also listed in this table. The physicochemical properties of the produced biodiesel were measured and benchmarked against the biodiesel standards based on ASTM D6751 and EN14214. It appears that all physicochemical properties of PME are sufficient to meet the ASTM and EN biodiesel standards. In particularly, the kinematic viscosity of the transesterified palm oil was substantially improved, but it was slightly higher than petroleum diesel. In addition, the resultant flash points for PME were relatively higher in compared with petroleum diesel and were suitable for use as a transportation fuel. However, the calorific value of the PME was lower than that of petroleum diesel. Another key property that significantly influences engine performance, emissions, and combustion characteristics is the cetane number of fuel. PME has a higher cetane number than petroleum diesel fuels.

Table 1 The fuel properties of petroleum diesel, PME biodiesel and gasoline

Properties	Unit	Diesel fuel	Biodiesel		PME	Gasoline
			Limit (ASTM D6751)	Test method
Kinematic viscosity @ 40 °C	mm^2 s^−1	3.34	1.9–6.0	D445	4.4	0.567
Density @ 15 °C	kg m^−3	851.9	880	D127	882.5	0.745
Acid number	mg KOH g^−1	0.12	0.5 max	D664	0.2	—
Calorific value	MJ kg^−1	45.31	—	D240	39.98	43.5
Flash point	°C	71.5	130 min	D93	155.5	−36
Pour point	°C	1	Not specified	D2500	4	—
Cloud point	°C	8	Not specified	D2500	5	—
Oxidation stability @ 100 °C	hours	>40	3 min	EN14112	15	—
Cetane number	—	52	47 min	D6890	75	—

---

In this study, all experiments were conducted under a constant speed of 1500 rpm and injection pressure of 600 bar. Two kinds of dual-fuel experiments, i.e. the DI diesel coupled with PFI gasoline and the DI PME fuel coupled with PFI gasoline, are compared in terms of performance, emissions, and combustion characteristics. Gasoline was port fuel injected onto the opened intake valve at −360° ATDC. Experiments were performed at five EGRs, 30, 35, 40, 45 and 50%. At each EGR level, SOI timing was varied from −5° ATDC and advanced up to the point at which potential unstable combustion starts to occur. For each type of DI fuel, the injection quantity was set to 6.5 mg per stroke for baseline diesel fuel and 7.6 mg per stroke for PME fuel, respectively. Considering the lower calorific value of PME fuel compared to baseline diesel, higher injection quantity is necessary to ensure equivalent fuel energy is injected for every cycle. Due to the introduction of dual fuels, a parameter, R_g, represents the ratio of energy of the premixed gasoline fuel Q_g to the total energy Q_t, which can be obtained from the following equation:

where m_g is the mass of the premixed gasoline fuel, m_d is the mass of the directly injected fuel, h_u is the calorific value and subscripts g and d denote premixed and directly injected fuel, respectively. In this study, the gasoline ratio was maintained at 0.6 for both direct injected diesel and PME dual-fuel combustion. In addition, the total supplied fuel energy is approximately 760 J per cycle.

The operating condition for this injection strategy is shown in Table 2. In each test, diesel fuel was used as the baseline fuel for comparison. When the engine was fuelled with biodiesel fuel, the engine ran satisfactorily throughout the entire test, which was performed at room temperature, and had no starting difficulties. The tests were performed under steady-state conditions with a sufficiently warmed exhaust gas and water coolant temperature. To enhance the accuracy of the study, each test point was repeated twice to produce average readings. The repeatability was matched over 95% for each test.

Table 2 Experimental conditions

Condition	Diesel/gasoline	PME/gasoline
Engine speed (rpm)	1500	1500
D.I. rail pressure (bar)	600	600
D.I. timing (° CA ATDC)	−5 to −95	−5 to −95
EGR rate (%)	30	30
D.I. fuel type	Diesel	PME
D.I. fuel quantity (mg per stroke)	6.5	7.6
P.I. fuel type	Gasoline	Gasoline
P.I. fuel quantity (mg per stroke)	10.4	10.4
Gasoline ratio (Rg)	0.6	0.6
Constant total equivalent fuel energy (J) per cycle	760	760

---

2.2. Engine and fuel delivery system

The test engine used in this study is based on a modified single-cylinder compression ignition diesel engine. The related engine modifications of the fuel delivery system of this engine have been discussed in the author's previous work.⁴⁷ The specifications of the test engine are listed in Table 3, and the schematic diagram of the experiment setup is shown in Fig. 1. The gasoline/diesel or gasoline/biodiesel dual-fuel engine experiments were performed using a port fuel injection of gasoline and direct injection of diesel/biodiesel. In dual fuel operation mode, the gasoline fuel was injected into the intake port just upstream of the intake valve, and the injection was timed to coincide with the cylinder's intake stroke. The gasoline fuel injector is based on an automotive style port fuel injector (Delphi), and the fuel line pressure was maintained at 400 kPa using an automotive pressure regulator. The injections and engine controls were managed using a programmable microcontroller and interface with Labview software. The microcontroller featured a programmable peak and hold pulse-width-modulation (PWM) to effectively drive the solenoid injectors for common-rail direct injection and port fuel injection separately. The control unit was designed to fully support and control engine parameters. The control unit was programmed to allow adjustment in various key engine operation parameters, including start of injection (SOI) timing, injection quantity, number of injections per cycle (pilot, main and post), injection pressure for DI injector, and injection timing and duration for port injector. The same controller system was able to simultaneously control the exhaust gas recirculation (EGR) system. The EGR was adopted to moderate the heat release rate (HRR) and combustion timing phasing. In particular, this involved the installation of EGR valve, EGR cooler, EGR surge tank, and two identical CO₂ sensors. The EGR rate can be flexibly adjusted by controlling the EGR valve. Under steady-state conditions, the EGR rate can be measured by comparing the ratio of CO₂ level in the intake to the exhaust as follows:

Table 3 Characteristics of single-cylinder engine

Parameter		Units
Displacement	638	cm^3
Bore	92	mm
Stroke	96	mm
Compression ratio	17.7 : 1
Rated power	7.8	kW
Rated speed	2400	rpm
D/Hbowl	2.81
Combustion chamber	Re-entrant type

---

Fig. 1 Schematic diagram of the experiment setup.

2.3. Instrumentation

The engine load absorber is based on the 7.5 kW A.C. synchronous dynamometer. It is used to provide loading to the engine and to maintain the engine speed. An airflow metre turbine with 2 to 70 litres per second (L s⁻¹) measuring range was installed to measure the intake airflow rate. To monitor the exhaust gas temperature, a type K thermocouple was mounted in the exhaust stream. The fuel flow rate for direct injection and the port fuel injection system were measured separately with a positive displacement gear wheel flow metre, which interfaced with a flow rate totalizer. The test system was installed with the necessary sensors for combustion analysis and fuel injection timing identification. In-cylinder gas pressure was measured using a Kistler 6125B type pressure sensor. The charge signal output of the pressure sensor was converted to a low-impedance voltage signal using a PCB model 422E53 in-line charge converter powered by a PCB model 480E09 signal conditioner. To acquire the top dead centre (TDC) position and crank angle signal for every engine rotation, an incremental quadrature rotary shaft angle encoder with 0.125° CA resolution (X4 encoding) was used. To determine and verify the SOI timing and injection duration for both of the injectors, the injector current signal was measured with a hall effect current sensor. To simultaneously acquire the cylinder pressure signal, injector current signal, and encoder signals, a computer equipped with a high-speed simultaneous sampling data acquisition card, which has 14 bit resolution, 2 MS s⁻¹ sampling rate, and four analog input channels, was used. The acquired data were further processed and analysed with Matlab software. In each test, 100 consecutive combustion cycles of pressure data were collected, and an average was calculated. To reduce noise effects, smooth data using SPAN as the number of points used to compute each element was applied to the sampled cylinder pressure data. Combustion parameters such as peak pressure magnitude, peak pressure location, heat release rate, peak heat release rate location, and ID were all computed using Matlab software. For the exhaust emission measurement, an AVL DICOM 4000 5-gas analyser was used to measure the concentrations of HC, CO, CO₂, and NO_x. Opacity of smoke was measured using AVL DiSmoke 4000. All emissions were measured during steady-state engine operation. The measurement range and resolution of both the instruments are provided in Table 4. The HC, CO, and NO_x emissions were converted into brake specific emissions by using the following equations according to SAE J177:

BSHC_{(g kW⁻¹ h⁻¹)} = 0.0287 × HC_(ppm) × exhaust mass flow rate_{(kg min⁻¹)}/brake power_(kW) (3)

BSCO_{(g kW⁻¹ h⁻¹)} = 0.0580 × CO_(ppm) × exhaust mass flow rate_{(kg min⁻¹)}/brake power_(kW) (4)

BSNO_{x(g kW⁻¹ h⁻¹)} = 0.0952 × NO_x(ppm) × exhaust mass flow rate_{(kg min⁻¹)}/brake power_(kW) (5)

Table 4 Measuring components, ranges and resolution of the AVL DICOM 4000 gas analyzer and DiSmoke 4000 smoke analyzer

Equipment	Measurement principle	Component	Measurement range	Resolution
Gas analyzer	Non-dispersive infrared	Unburned hydrocarbon (HC)	0–20 000 ppm	1 ppm
	Non-dispersive infrared	Carbon monoxide (CO)	0–10% vol	0.01% vol
	Non-dispersive infrared	Carbon dioxide (CO2)	0–20% vol	0.1% vol
	Electrochemical	Nitrogen oxides (NOx)	0–5000 ppm	1 ppm
	Calculation	Excess air ratio (λ)	0–9999	0.001
Smoke opacimeter	Photodiode detector	Opacity (%)	0–100%	0.10%

---

3. Calculation methods

3.1. Engine performance

Engine performance in this work was evaluated based on BSFC and BTE, which were determined and calculated according to the following equations:

3.2. Combustion analysis

HRR analysis is a useful approach for assessing the effects of fuel injection system, fuel type, engine design changes, and engine operating conditions on the combustion process and engine performance.⁴⁸ Given the plot of HRR versus crank angle, it is easy to identify the start of combustion (SOC) timing, the fraction of fuel burned in the premixed mode, and differences in combustion rates of fuels.⁴⁹ In this paper, different type of fuels were fuelled in an identical compression ignition engine; hence, the HRR information is an important parameter for interpreting engine performance and exhaust emissions. In this study, the average in-cylinder pressure data of 100 successive cycles, acquired with a 0.125° crank angle resolution, were used to compute the HRR. The HRR, given by , at each crank angle was obtained from the first law of thermodynamics, and it can be calculated by the following formula:where, γ = specific heat ratio, P = instantaneous cylinder pressure (Pa), and V = instantaneous cylinder volume (m³).

3.3. Statistical and equipment uncertainty analysis

Experimental errors and uncertainties can arise from instrument selection, condition, calibration, environment, observation, reading, and test procedure. The measurement range, accuracy, and percentage uncertainties associated with the instruments used in this experiment are listed in Table 5. Uncertainty analysis is necessary to verify the accuracy of the experiments. Percentage uncertainties of various parameters such as brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), brake specific hydrocarbon (BSHC), brake specific carbon monoxide (BSCO), and brake specific nitrogen oxide (BSNO_x) were determined using the percentage uncertainties of various instruments employed in the experiment. To compute the overall percentage uncertainty from the combined effect of the uncertainties of various variables, the principle of propagation of errors can be estimated as ±4.3%. The overall experimental uncertainty was computed as follows:

Overall experimental uncertainty = square root of [(uncertainty of fuel flow rate)² + (uncertainty of BSFC)² + (uncertainty of BTE)² + (uncertainty of BSCO)² + (uncertainty of BSNO_x)² + (uncertainty of EGT)² + (uncertainty of smoke)² + (uncertainty of pressure sensor)² + (uncertainty of crank angle encoder)²] = square root of [(2)² + (1.95)² + (1.74)² + (2.22)² + (0.73)² + (0.15)² + (1)² + (1)² + (0.03)²] = ±4.3%

Table 5 List of measurement accuracy and percentage uncertainties

Measurement	Measurement range	Accuracy	Measurement techniques	% Uncertainty
Load	±120 N m	±0.1 N m	Strain gauge type load cell	±1
Speed	60–10 000 rpm	±1 rpm	Magnetic pick up type	±0.1
Time	—	±0.1 s	—	±0.2
Fuel flow measurement	0.5–36 L h^−1	±0.01 L h^−1	Positive displacement gear wheel flow meter	±2
Air flow measurement	2–70 L s^−1	±0.04 L s^−1	Turbine flow meter	±0.5
CO	0–10% by vol.	±0.001%	Non-dispersive infrared	±1
NOx	0–5000 ppm	±1 ppm	Electrochemical	±1.3
Smoke	0–100%	±0.1%	Photodiode detector	±1
EGT sensor	0–1200 °C	±0.3 °C	Type K thermocouple	±0.15
Pressure sensor	0–25 000 kPa	±12.5 kPa	Piezoelectric crystal type	±1
Crank angle encoder	0–12 000 rpm	±0.125°	Incremental optical encoder	±0.03
Computed
BSFC	—	±7.8 g kW^−1 h^−1	—	±1.95
BTE	—	±0.5%	—	±1.74
BSCO	—	±0.1 g kW^−1 h^−1	—	±2.22
BSNOx	—	±0.1 g kW^−1 h^−1	—	±0.73

---

4. Results and discussions

4.1. SOI timing sweep at constant EGR

4.1.1. Performance analysis. The indicated specific fuel consumption (ISFC) as a function of SOI timing for baseline diesel/gasoline and PME/gasoline dual-fuel experiments is shown in Fig. 2. In general, the results indicated that the ISFC increases with an advance of SOI timing. Furthermore, the ISFC for PME/gasoline operation is constantly greater than that of baseline diesel/gasoline across all SOI timings. The higher ISFC of PME/gasoline corresponds to less efficient operation, thus requiring a greater amount of DI fuel to accomplish the same amount of indicated power. This is due to the low calorific value of PME compared with diesel, which was about 12% lower than that of baseline diesel fuel. Subsequently, the results also indicate that the ISFC tends to be highly sensitive to variation in SOI timing. For both the dual fuel operations, advancing the SOI timing resulted in an increase in ISFC initially that reached the highest value at SOI timing of −45° ATDC and then was reduced. With advancing SOI, a large fraction of fuel burns in premixed combustion phase causes earlier occurrence of high peak combustion pressure that potentially resists upward movement of the piston. However, with SOI earlier than −45° ATDC, the fuel air became more homogeneously mixed, and the low equivalence ratio extended the ignition delay of the directly injected fuel and retarded the combustion phasing.⁵⁰ This is the turning point at which ISFC began to drop off with further SOI advancement. However, further SOI advance beyond −95° ATDC caused unstable combustion, and the results are not presented here. With 30% EGR rate, the lowest achievable ISFCs for diesel/gasoline and PME/gasoline operation are found to be 187.7 and 203.1 g kW⁻¹ h⁻¹, respectively, at an optimum SOI timing of −7° ATDC.

Fig. 2 ISFC at various SOI timing for dual fuel operation diesel/gasoline and PME/gasoline fuels, at 30% EGR.

Fig. 3 illustrates the variations in BTE with different SOI timings for engines fuelled with diesel/gasoline and PME/gasoline dual fuel combustion at 30% EGR. The PME/gasoline operation shows slightly higher BTE than diesel/gasoline with SOI between −5 to −15° ATDC, and somewhat lower with further SOI advancement. The highest reported BTE for diesel/gasoline and PME/gasoline are found to be 34.8% and 35.5%, respectively, at SOI timing of −9° ATDC. In both the dual fuel operations, advancing the SOI timing resulted in an increase in BTE initially, reached the peak value at SOI timing of −9° ATDC, and then was reduced. Some improvement in BTE was gained with SOI beyond −50° ATDC. The increment effect is because the combustion phasing retards as SOI advances. This phenomenon also suggests that the extended mixing time of the advanced SOI timings allows the DI fuel to penetrate more thoroughly throughout the combustion chamber. As a result, the local fuel reactivity of the most reactive regions in the combustion chamber is reduced, and the combustion phasing is delayed.⁵⁰ However, shifting the SOI to −95° ATDC caused the BTE to drop off sharply. The excessive advance of SOI timing tends to lean out the direct injected fuel distribution in the cylinder at SOC, which did not permit the combustion progression and caused the engine to begin running rough with intermittent misfiring.³⁴ Thus, further enleanment (i.e., advanced SOI) will only result in worse emissions and performance

Fig. 3 BTE at various SOI timing for dual fuel operation of diesel/gasoline and PME/gasoline fuels, at 30% EGR.

4.1.2. Emissions analysis. The variation of NO_x emissions for both the dual fuel operations under various SOI timings and at 30% EGR is illustrated in Fig. 4. Generally, the PME/gasoline operation shows slightly lower NO_x emissions than diesel/gasoline across all SOI timings. In fact, both fuels operation behave the same in terms of NO_x variation as a function of SOI timing. With advancing SOI, NO_x is initially increased and reaches the maximum value of 9.8 and 9.7 for diesel/gasoline and PME/gasoline, respectively, at SOI −35° ATDC, and then is reduced. The results suggest that too advanced or too retarded SOI timing will both reduce NO_x. A substantially lower level of NO_x below EURO VI emission standards can be achieved by early or late SOI timing for both fuel operations. With advanced SOI cases, the early injection tend to lean out the local equivalence ratio resulting from extended mixing time. The effect of reduced local equivalence ratios retards the combustion phasing, resulting in lower peak flame temperatures and therefore lower NO_x emissions.⁵⁰ On the other hand, with late injection cases, the combustion phasing is progressively retarded and shifted away from TDC in expansion stroke. This effect leads to reduced combustion temperature and lower NO_x formation via thermal or Zeldovich mechanism. The effect of SOI timing on combustion phasing can be realised from the CA50 plot as shown in Fig. 5. As can be seen, later or earlier SOI timing will all shift the CA50 away from TDC in the expansion stroke, which explains the decreased NO_x emissions. A minimum point of CA50 versus SOI timing presents for both cases. This phenomenon has been reported in other investigations of dual fuel combustion in diesel engine.⁵¹ In addition, it is interesting to note that the same CA50 is possibly obtained by early or late SOI timing. With advanced SOI timing, the leaner local equivalence ratios prolonged the ignition delay and thus caused ignition to occur late. However, when SOI was retarded, ignition delay became shorter due to the considerably higher cylinder pressure and temperature at that moment, which led to immediate combustion after DI fuel was injected.

Fig. 4 BSNO_x emissions at various SOI timing for dual fuel operation of diesel/gasoline and PME/gasoline fuels, at 30% EGR.

Fig. 5 CA50 at various SOI timing for dual fuel operation of diesel/gasoline and PME/gasoline fuels, at 30% EGR.

Fig. 6 shows the variations in HC emissions with different SOI timings for engine fuelled with diesel/gasoline and PME/gasoline dual fuel combustion at 30% EGR. The PME/gasoline operation reveals lower HC than diesel/gasoline across all SOI timings. The oxygen content in PME fuel is especially useful in limiting locally fuel rich regions, resulting in improved combustion and thereby reducing HC emissions. Furthermore, it can be seen that the highest HC emissions resulting from diesel/gasoline and PME/gasoline operation are 2.14 and 2.07 g kW⁻¹ h⁻¹, respectively, with earliest SOI timing of −95° ATDC. As previously discussed, the earliest SOI timing tends to lean out the overall fuel mixture and caused the engine to begin running rough with intermittent misfiring, thereby resulting in higher HC emissions. Some HC increments were observed with SOI advancement between −5° to −17° ATDC. However, a reducing trend in HC emissions was observed with further SOI advancement from −19° to −55° ATDC. The effects of HC reduction with advanced SOI timing have suggested that there was continuous improvement in DI fuel distribution that permits the premixed gasoline in the squish area to burn more completely, thereby decreasing HC emissions. Fig. 7 shows the effect of SOI timing on CO emissions for engines fuelled with diesel/gasoline and PME/gasoline dual fuel combustion at 30% EGR. Similarly to the HC variation trend, it can be seen that CO emissions first decrease with advancing SOI timing, reaching a maximum at an SOI of −17° ATDC, and then showing a gradual decrease. However, with considerably advanced SOI timing of beyond −65° ATDC, a steep increase in CO emissions was observed. In fact, it can be seen that CO emissions were lower for PME/gasoline than diesel/gasoline across all SOI timings. The use of oxygenated fuel of PME biodiesel would be expected to enhance combustion efficiency, thereby reducing CO emissions.

Fig. 6 BSHC emissions versus SOI timing sweeps at 30% EGR for dual fuel operation of diesel/gasoline and PME/gasoline fuels.

Fig. 7 BSCO emissions versus SOI timing sweeps at 30% EGR for dual fuel operation of diesel/gasoline and PME/gasoline fuels.

4.1.3. Combustion analysis. A more comprehensive study of combustion characteristics resulting from SOI timing variation for both the dual-fuel operation at 30% EGR is shown in Fig. 8. At advanced SOI timing, peak cylinder pressure is initially increased and reaches the maximum at SOI timing of −35° ATDC, and then is reduced. This indicates that too early or too late SOI timing will both reduce the peak cylinder pressure and cause delay in combustion phasing. It can be seen that at different SOI timings, the peak cylinder pressure and peak HRR of the PME/gasoline are marginally higher than those of diesel/gasoline cases except for the considerably advanced SOI timing past −35° ATDC. Compared to the case of SOI of −35° ATDC, both late and early SOI timings will reduce the peak cylinder pressure and retard the corresponding crank angle toward the expansion stroke. This interesting effect can be explained because as the SOI advanced, the in-cylinder mixture became more homogenous, and the low equivalence ratio extended the ignition delay of DI fuel, thereby delaying the combustion phasing. For retarded SOI cases, on the other hand, the combustion process becomes more coupled with the variation in SOI timing. From the HRR profile, which was calculated from in-cylinder pressure, it can be observed clearly that a remarkable two-stage high temperature heat release (HTHR) occurred for late SOI timing of −7° to −15° ATDC. In contrast, for advanced SOI timing cases, the combustion process is characterised by single-stage low temperature heat release (LTHR) and followed by single-stage HTHR. The first stage is called cool flame reaction, and it proceeds at temperatures below the auto-ignition temperature of the fuel, as explained by Pekalski et al.⁵² In this investigation, this reaction appears consistently as a small peak of HRR at approximately −17° ATDC. This peak reflects low temperature reactions (LTR), and the corresponding crank angle is nearly at the same position of crank angle regardless of the SOI timing. In the LTR stage, part of the premixture of gasoline and injected DI fuel is consumed through an initial breakdown of fuel molecules, leading to the formation of free radicals, aldehydes, and hydrogen peroxide. Because of the heat released in the LTR stage, the mixture temperature rises and causes the remainder mixture to combust, leading to another stage of HTHR combustion. Meanwhile, it can be seen that the peak and the timing of this second stage HTHR are influenced by varying the SOI timing. As the SOI is advanced from −35 to −55° ATDC, the peak of second stage HTHR increases and becomes narrower. However, further advancing the SOI past −55° ATDC caused a reduction in this peak. With SOI advancement, more time is being provided for the DI fuel to mix with the premixture of gasoline and air; thus, the mixture becomes more uniform, and the local equivalence ratio decreases. The leaner local equivalence ratio prolonged the ignition delay and retarded combustion events toward TDC in the compression stroke, thus increasing the peak of second stage HTHR. Because the combustion proceeded without flame propagation, resulting in lower locally combustion temperature, thereby can be used to explain the reduction in NO_x emissions. Further advancing the SOI past −55° ATDC increased the time for the preparation of a highly homogenous mixture and kept reducing the local equivalence ratio, thus further extending the ignition delay and retarding combustion phasing toward the expansion stroke. The extended ignition delay retards combustion phasing, leading to a larger fraction of heat being released near TDC in the expansion stroke and causing a decrease in in-cylinder combustion temperature. Consequently, the NO_x emissions tend to be further reduced. In addition, in general, under most of the SOI timing conditions except for SOI of −75° ATDC, all PME/gasoline operations exhibited advanced SOC timing than the diesel/gasoline operations owing to their relatively higher cetane number. An additional result on the COV (coefficient of variance) of IMEP (indicated mean effective pressure) as shown in Fig. 9 is often used to indicate combustion stability. In general, better combustion stability corresponds to a lower COV value. The results indicate that advancing the SOI timing past −25° ATDC increased the cyclic variations almost linearly, and they continued to increase rapidly with the earliest SOI. Cyclic variation in in-cylinder combustion pressure over 100 consecutive cycles for both of diesel/gasoline and PME/gasoline dual fuel operations at four selected SOI timings are shown in Fig. 10. Practically speaking, cyclic variation is undesirable because it worsens overall engine efficiency, performance, and emissions. From the results, the cause of cyclic variation (COVimep) with the shift in SOI timing can be clearly observed. When SOI advanced beyond −35° ATDC, an obvious inverse correlation between combustion phasing and SOI timing is noticed from the previously presented HRR results. That is the point at which the cylinder mixture equivalence ratio begins to dominate the combustion stability. For both of the dual fuel operations, advancing the SOI timing resulted in a decrease in COVimep initially, and it reached the lowest value at SOI timing of −15° ATDC and increased. With the considerably advanced SOI of −85° ATDC, the largest portion of the mixture is at the lowest equivalence ratio, and the extended ignition delay retarded the combustion phasing. The late combustion burned under considerably low temperature and caused incomplete combustion, therefore leading to higher cycle-to-cycle variation. Further advancing in SOI timing to −95° ATDC worsened the combustion stability and increased the probability of misfiring. Compared to diesel/gasoline operation, the PME/gasoline operation exhibited higher COV of IMEP with SOI past −25° ATDC. The oxygen content in the PME fuel produced more mixture with much leaner equivalence ratio, thus causing an increase in cyclic variation.

Fig. 8 Effect of SOI timing on combustion pressure and heat release rate for dual fuel operation of diesel/gasoline and PME/gasoline at 30% EGR.

Fig. 9 Effect of SOI timing variation on the coefficient of variation of indicated mean effective pressure for dual fuel operation of diesel/gasoline and PME/gasoline at 30% EGR.

Fig. 10 Comparison of 100 cycles of cylinder pressure between diesel/gasoline and PME/gasoline dual fuel operation at EGR = 30%.

4.2. Dual-fuel with EGR sweep

The previous section revealed how combustion processes proceeded differently as the SOI timing was changed under a constant EGR rate. In-cylinder charge reactivity and distribution can also be affected by the effect of EGR variation. To investigate the dilution effect of EGR on engine performance, emissions, and combustion characteristics for diesel/gasoline and PME/gasoline dual-fuel operation, five EGR levels of 30%, 35%, 40%, 45% and 50% are examined. Fig. 11 shows the results. With increasing EGR rate, it is feasible to maintain ISFC at the minimum possible magnitude by advancing SOI timing for diesel/gasoline and PME/gasoline operation. This is due to the effect of combustion phasing retardation as EGR increases; thus, the cooperative control of SOI timing adjustment can be used to compensate for the drop in combustion efficiency. At a higher EGR rate, the SOI timing for the diesel/gasoline operation is earlier than for the PME/gasoline operation, suggesting that higher cetane number and shorter ignition delay of the directly injected PME fuel improve the combustion of the fuel mixture. Furthermore, it can be seen that the maximum pressure rise rate (MPRR) decreased with increasing EGR rate for both cases. This indicates the advantage of EGR in controlling and regulating the HRR, thus lowering the MPRR. In terms of emissions characteristics, HC and CO increase with increasing EGR rate due to the reduced in-cylinder gas temperature with dilution and increased incomplete combustion. Conversely, smoke and NO_x decrease with the increasing dilution effect of EGR. With higher EGR, the prolonged ignition delay has resulted in the formation of more uniform mixtures that burn at substantially lower temperatures, which leads to considerably lower NO_x emissions (below EURO VI limit of 0.4 g kW⁻¹ h⁻¹).

Fig. 11 Influence of EGR variation on performance, emissions and combustion characteristics for engine operation with dual fuel combustion of diesel/gasoline and PME/gasoline.

5. Conclusions

The main objective of this research is to use biofuels as alternative energy sources for engines operating in RCCI dual-fuel combustion mode. A series of orderly experiments was conducted to evaluate the use of biodiesel fuels in internal combustion engines operating with dual-fuel combustion strategies. The assessment of engine operation under RCCI dual-fuel combustion mode was carried out using diesel or biodiesel as direct injected fuel and gasoline as port injected fuel. Based on the experimental results, the following main conclusions can be drawn from this investigation.

(1) With 30% EGR rate, the lowest achievable ISFCs for diesel/gasoline and PME/gasoline operation are found to be 187.7 and 203.1 g kW⁻¹ h⁻¹, respectively, at an optimum SOI timing of −7° ATDC.

(2) The PME/gasoline operation shows slightly higher BTE than diesel/gasoline with SOI between −5 to −15° ATDC, and somewhat lower with further SOI advancement. The highest reported BTE for diesel/gasoline and PME/gasoline are found to be 34.8% and 35.5%, respectively, at SOI timing of −9° ATDC.

(3) A substantially lower level of NO_x below EURO VI emission standards can be achieved by early or late SOI timing for both fuel operations. Besides, the same CA50 is possibly obtained by early or late SOI timing.

(4) The PME/gasoline operation reveals lower HC and CO than diesel/gasoline across all SOI timings. The oxygen content in PME fuel is especially useful in limiting locally fuel rich regions, resulting in improved combustion and thereby reducing HC and CO emissions.

(5) Compared to diesel/gasoline operation, the PME/gasoline operation exhibited higher COV of IMEP with SOI past −25° ATDC. The oxygen content in the PME fuel produced more mixture with much leaner equivalence ratio, thus causing an increase in cyclic variation.

(6) With increasing EGR rate, it is feasible to maintain ISFC at the minimum possible magnitude by advancing SOI timing for diesel/gasoline and PME/gasoline operation. In addition, the prolonged ignition delay under higher EGR level has resulted in the formation of more uniform mixtures that burn at substantially lower temperatures, which leads to considerably lower NO_x emissions (below EURO VI limit of 0.4 g kW⁻¹ h⁻¹).

In conclusion, the results from this study suggest that alternative fuels from bio resources have high potential as substitutes for petroleum-based fuels for engines operating with low temperature combustion strategies. Meanwhile, the engine operating in RCCI dual-fuel combustion mode is capable of achieving high efficiency with near zero NO_x and smoke emissions.

Nomenclature and symbol

ASTM	American society for testing and materials
ATDC	After top dead centre
BMEP	Brake mean effective pressure
BSCO	Brake specific carbon monoxide
BSFC	Brake specific fuel consumption
BSHC	Brake specific hydrocarbon
BSNOx	Brake specific nitrogen oxide
BTE	Brake thermal efficiency
CA	Crank angle
CA50	Burn point of 50%
CAI	Controlled auto-ignition
CI	Compression ignition
CO	Carbon monoxide
CO2	Carbon dioxide
COV	Coefficient of variance
DI	Direct injection
ECU	Electronic control unit
EGR	Exhaust gas recirculation
FAME	Fatty acid methyl ester
HC	Hydrocarbon
HCCI	Homogenous charge compression ignition
HCDC	Homogeneous charge diesel combustion
HCLI	Homogeneous charge late injection
HRR	Heat release rate
HTHR	High temperature heat release
IMEP	Indicated mean effective pressure
ISFC	Indicated specific fuel consumption
LTC	Low temperature combustion
LTHR	Low temperature heat release
LTR	Low temperature reactions
MK	Modulated kinetics
MPRR	Maximum pressure rise rate
NOx	Nitrogen oxide
PCCI	Premixed charge compression ignition
PFI	Port fuel injection
PME	Palm methyl ester
PWM	Pulse-width-modulation
RCCI	Reactivity controlled compression ignition
Rpm	Revolution per minute
SI	Spark ignition
SOC	Start of combustion
SOI	Start of injection
TDC	Top dead centre
UNIBUS	Uniform bulky combustion system
λ	Relative air-fuel ratio (lambda)

Acknowledgements

The authors would like to acknowledge the Ministry of Higher Education (MOHE) of Malaysia and University of Malaya for financial support through HIR grant (UM.C/HIR/MOHE/ENG/07) and Postgraduate Research Grant (PPP) (grant number PG035-2012B).

References

1. M. Mofijur, H. H. Masjuki, M. A. Kalam, A. E. Atabani, M. Shahabuddin, S. M. Palash and M. A. Hazrat, Renewable Sustainable Energy Rev., 2013, 28, 441–455.
2. A. M. Liaquat, M. A. Kalam, H. H. Masjuki and M. H. Jayed, Atmos. Environ., 2010, 44, 3869–3877.
3. B. Tesfa, R. Mishra, F. Gu and A. D. Ball, Renewable Energy, 2012, 37, 333–344.
4. K. Anand, R. P. Sharma and P. S. Mehta, Biomass Bioenergy, 2011, 35, 533–541.
5. A. K. Agarwal, Prog. Energy Combust. Sci., 2007, 33, 233–271.
6. M. Mofijur, H. H. Masjuki, M. A. Kalam, M. A. Hazrat, A. M. Liaquat, M. Shahabuddin and M. Varman, Renewable Sustainable Energy Rev., 2012, 16, 5007–5020.
7. J. E. Dec, Proc. Combust. Inst., 2009, 32, 2727–2742.
8. W. F. Northrop, D. N. Assanis and S. Bohac, SAE Int. J. Engines, 2011, 4, 1431–1444.
9. Y.-Y. Wu, C.-T. Jang and B.-L. Chen, J. Eng. Gas Turbines Power, 2010, 132, 044501–044504.
10. I. Lemberger and G. Floweday, SAE Int. J. Engines, 2009, 2, 1559–1573.
11. N. Shimazaki, T. Tsurushima and T. Nishimura, SAE International, 2003-01-0742, 2003.
12. K. Okude, K. Mori, S. Shiino and T. Moriya, SAE International, 2004-01-1907, 2004.
13. M. P. B. Musculus, S. Singh and R. D. Reitz, Combust. Flame, 2008, 153, 216–227.
14. K. K. Srinivasan, S. R. Krishnan, Y. Qi, K. C. Midkiff and H. Yang, Combust. Sci. Technol., 2007, 179, 1737–1776.
15. S. Singh, R. D. Reitz, M. P. B. Musculus and T. Lachaux, Combust. Sci. Technol., 2007, 179, 2381–2414.
16. A. K. Agarwal, SAE International, 2011-01-1183, 2011.
17. T. Urushihara, K. Hiraya, A. Kakuhou and T. Itoh, SAE International, 2003-01-0749, 2003.
18. W. Su, T. Lin and Y. Pei, SAE International, 2003-01-0741, 2003.
19. S. Juttu, S. S. Thipse, N. V. Marathe and M. K. Gajendra Babu, SAE International, 2007-26-020, 2007.
20. S. Gan, H. K. Ng and K. M. Pang, Appl. Energy, 2011, 88, 559–567.
21. X. Han, M. Wang and M. Zheng, SAE International, 2015-01-0001, 2015.
22. H. Bendu and S. Murugan, Renewable Sustainable Energy Rev., 2014, 38, 732–746.
23. R. H. Thring, SAE International, 892068, 1989.
24. R. Hasegawa and H. Yanagihara, SAE International, 2003-01-0745, 2003.
25. S. Kimura, O. Aoki, H. Ogawa, S. Muranaka and Y. Enomoto, SAE International, 1999-01-3681, 1999.
26. M. Jia, Y. Li, M. Xie, T. Wang, H. Wang and R. D. Reitz, Energy Procedia, 2015, 66, 33–36.
27. A. J. Torregrosa, A. Broatch, A. García and L. F. Mónico, Appl. Energy, 2013, 104, 149–157.
28. T. Aoyama, Y. Hattori, J. i. Mizuta and Y. Sato, 1996.
29. M. Himabindu and N. V. Mahalakshmi, SAE International, 2007-01-4065, 2007.
30. M. Himabindu, G. Sathyanarayanan, S. Thomson and N. V. Mahalakshmi, SAE International, 2008-01-2609, 2008.
31. M. Odaka, H. Suzuki, N. Koike and H. Ishii, SAE International, 1999-01-0184, 1999.
32. D. J. Cook, H. Pitsch and G. Nentwig, SAE International, 2008-01-1666, 2008.
33. J. Li, W. M. Yang, H. An and D. Zhao, Appl. Energy, 2015, 155, 59–67.
34. J. Benajes, S. Molina, A. García and J. Monsalve-Serrano, Energy Convers. Manage., 2015, 99, 193–209.
35. A. Bhave, M. Kraft, F. Mauss, A. Oakley and H. Zhao, SAE International, 2005-01-0161, 2005.
36. I. W. Ekoto, W. F. Colban, P. C. Miles, S. W. Park, D. E. Foster, R. D. Reitz, U. Aronsson and Ö. Andersson, SAE Int. J. Engines, 2009, 2, 411–430.
37. P. Das, P. M. V. Subbarao and J. P. Subrahmanyam, SAE International, 2014-01-2661, 2014.
38. S. L. Kokjohn, R. M. Hanson, D. A. Splitter and R. D. Reitz, SAE Int. J. Engines, 2009, 2, 24–39.
39. R. D. Reitz and G. Duraisamy, Prog. Energy Combust. Sci., 2015, 46, 12–71.
40. R. D. Reitz, Proceedings of the Directions in Engine-Efficiency and Emissions Research (DEER) Conference Presentations, Detroit, MI, 27–30 September, 2010.
41. M. F. Al-Dawody and S. K. Bhatti, Energy Convers. Manage., 2013, 68, 96–104.
42. S. M. Palash, H. H. Masjuki, M. A. Kalam, B. M. Masum, A. Sanjid and M. J. Abedin, Energy Convers. Manage., 2013, 76, 400–420.
43. S. M. A. Rahman, H. H. Masjuki, M. A. Kalam, M. J. Abedin, A. Sanjid and H. Sajjad, Energy Convers. Manage., 2013, 76, 362–367.
44. A.-H. Kakaee, P. Rahnama and A. Paykani, J. Nat. Gas Sci. Eng., 2015, 25, 58–65.
45. M. K. Veltman, P. K. Karra and S.-C. Kong, SAE International, 2009-01-0485, 2009.
46. B. Tompkins and T. Jacobs, Spring technical meeting of the central states section of the combustion institute, 2012, vol. 2.
47. Y. H. Teoh, H. H. Masjuki, M. A. Kalam, M. A. Amalina and H. G. How, RSC Adv., 2014, 4, 50739–50751.
48. J. Ghojel and D. Honnery, Appl. Therm. Eng., 2005, 25, 2072–2085.
49. M. Canakci, A. N. Ozsezen and A. Turkcan, Biomass Bioenergy, 2009, 33, 760–767.
50. S. Curran, V. Prikhodko, K. Cho, C. S. Sluder, J. Parks, R. Wagner, S. Kokjohn and R. D. Reitz, SAE International, 2010-01-2206, 2010.
51. W. Fang, B. Huang, D. B. Kittelson and W. F. Northrop, J. Eng. Gas Turbines Power, 2014, 136, 081502.
52. A. A. Pekalski, J. F. Zevenbergen, H. J. Pasman, S. M. Lemkowitz, A. E. Dahoe and B. Scarlett, J. Hazard. Mater., 2002, 93, 93–105.

---