Introduction

Malaysia is one of the leading exporters of palm oil and is able to produce its own palm oil biodiesel. However, as the European Union (EU) is becoming increasingly hostile towards palm oil imports1,2, Malaysia need to find ways to fully utilize this commodity in the domestic market. One of the most promising ways to increase palm oil utilization is by increasing palm oil-based biodiesel content in the domestic diesel. Since 2010, Malaysia have raised the percentage of biodiesel content in domestic diesel fuel from 5 to 10 vol%. The 12th Malaysia Plan (RMK-12) has set a goal to impose 30 vol% of biodiesel–diesel blend (B30) in domestic diesel fuel by the year 20253. As of now, commercially available diesel sold in service stations nationwide consist of either 7 vol% (B7) or 10 vol% (B10) biodiesel–diesel. Nonetheless, such efforts triggers concern among engine manufacturers as well as commercial and private vehicle owners as to how this implementation impacts fuel costs and efficiency, as well as engine performance and durability.

Biodiesel production cost is higher as compared to conventional petroleum-based diesel fuel4, therefore increments of biodiesel content in biodiesel–diesel fuel blends will result in rising fuel production costs. Domestically in Malaysia, the government has for decades imposed a large-scale subsidy program on fuel prices5 which may be able to absorb the price hike of increasing biodiesel–diesel fuel blends, thereby stabilizing the market price of commercial diesel. Nonetheless, as biodiesel content in biodiesel–diesel fuel increases, the overall calorific value per volume shall reduce due to biodiesels possessing lower calorific value (CV) than petroleum-based diesel. This shall lead to increased fuel consumption as more fuel shall be consumed in order to produce the same amount of energy as conventional petroleum-based diesel. Furthermore, fuel consumption could also increase as biodiesel possess higher viscosity which could lead to poor fuel pumping and spray behavior6.

Biodiesel in general possess higher kinematic viscosity and density than conventional diesel7. These factors affect fuel droplets atomization and entrainment when injected into the combustion chamber. However, in a modern diesel engine equipped with a common rail fuel injection system, it is argued that the effects of the aforementioned properties were rather insignificant. This is because, high pressure fuel injection introduced by the common rail fuel injection system enables improved fuel droplets atomization and evaporation, therefore enhancing combustion process8. Nonetheless, biodiesel emit higher nitrogen oxides (NOx) emissions when taking into account longer ignition delays caused by higher peak combustion temperatures as the result of higher oxygen content in biodiesel9. Furthermore, due to lower CV of biodiesel, fuel consumption is considerably higher compared to conventional diesel. Several research have documented higher fuel consumption when using various biodiesel blends in common rail injection diesel engines10,11,12.

In order to reduce NOx emissions and improve fuel consumption of biodiesel and biodiesel–diesel blends, there is a need to reduce in-cylinder temperatures and improve combustion efficiency. One strategy to achieve this is by using Water-in-diesel (W/D) emulsion fuel. The effects of W/D has been studied for many years and has shown promising improvements in terms of engine performance and exhaust emissions13,14,15,16. This can be attributed to micro-explosion of micro-sized water droplets dispersed in diesel oil. Micro-explosion occurs during combustion when water droplets suspended in W/D emulsions undergoes explosion due to it having lower boiling point. This results in a secondary atomization of the primary fuel spray, decreasing distribution of fuel droplets, resulting in improved combustion17,18. It was reported that as water content in W/D emulsion increases up to a maximum of 20%, Brake Thermal Efficiency (BTE) increases while gas temperature (EGT) behaved conversely, indicating lower peak temperatures due to the charge cooling by water evaporation15. In another study, soot emission was significantly by 50% by using W/D micro-emulsions containing 3.5 vol% of water19. In relation to biodiesel–diesel W/D emulsions, it was documented that W/D can reduce exhaust gases such as carbon monoxide (CO), unburned hydrocarbon (UHC), and soot opacity, while carbon dioxide (CO2) emissions increased20. The authors inferred that the CO2 increases observed are related to a more complete combustion, as well as hydroxide (OH) radicals present during water vaporization assists formation of CO2 from CO.

Nonetheless, since water and oil could not be mixed naturally, synthesis of W/D emulsions require the use of a chemical additive known as surfactant or emulsifier to suspend the water particles in the diesel oil for a sustained period of time. Despite its benefits, surfactants used in W/D emulsions are known to cause fuel filters clogging by displacing deposits in the fuel lines and fuel tanks21. Another shortcoming of surfactants is due to its expensive nature, large-scale production of W/D emulsion fuels would not be a viable substitute to petroleum-based diesel, rendering commercialization difficult22.

To remove the dependence on surfactants to produce W/D emulsion fuel, Ithnin et al.23 developed a device capable of producing W/D emulsion fuel without addition of any surfactant by incorporating a real time emulsifying device which mixes diesel and water within the fuel line and on-demand to the engine. It operates by using a high shear mixer in combination with an ultrasonic agitator to produce the W/D emulsion. This device was named Real-Time Non-Surfactant Emulsion Fuel Supply System; or in short, RTES. It was initially tested on a 5 kW single cylinder mechanical fuel injection diesel engine producing 6.5 wt% W/D emulsion fuel and the results proved that engine BTE improved by 3.59%, while Brake Specific Fuel Consumption (BSFC) was reduced by 3.89%. Exhaust emissions also showed favorable improvements, with NOx and particulate matter (PM) emissions plunged by 31.7% and 16.3% respectively when compared to conventional diesel.

Moving forward, further research has been done to examine the effects of RTES implementation on diesel engines. In general, RTES implementation was effective in lowering down NOx and smoke emissions with simultaneous increases in BTE as well as reductions in fuel consumption. A summary of previous research on the effects of W/D produced by RTES towards various engine applications are explained in Table 1.

Table 1 Summary of past research on the effects of W/D produced by RTES.
Full size table

However, one of the key elements of the original RTES design was the role of ultrasonic agitator as one of the mixing methods, which reduced overall energy efficiency due to its high-power requirement. To ensure smooth and stable supply of W/D emulsion to the engine, the ultrasonic agitator demands 120 W of electrical energy23. In addition, a study to determine RTES durability during extended use reported ultrasonic agitator failure after 26 h24.

In reaction to this, further design improvements were conducted RTES Technology (M) Sdn. Bhd. which eliminated the use of ultrasonic agitator from the mixing method25. The updated design consists of static mixers and booster pumps to facilitate fluid turbulence and promote mixing. This design concept was tested by Mahdi et al.26 and it was found that when 7–10 vol% of water content in W/D emulsion were mixed at 3500 rpm for 1 min, stability was maintained within 128 s. In the same study, pilot tests on a mechanical injection-type diesel engine showed lower NOx and fuel consumption, implying that W/D emulsions produced using this design exhibit similar qualities to the original RTES design.

This study is a continuation of our research on RTES-produced W/D emulsions as alternative fuel in industrial diesel-electric generators32. In this paper, the effect of B30 emulsion fuel (B30E) produced using the updated RTES design was evaluated. The main purpose of this study focuses on examining the effects of B30 emulsions with variable water contents towards lowering NOx emissions often associated with biodiesel–diesel blends. Secondly, since previous studies on RTES were conducted using only naturally aspirated mechanical-type fuel injection diesel engines with conventional diesel and/or low biodiesel–diesel blends (B10), it is important to establish the performance and emissions profile of a modern common rail injection-type engines using higher biodiesel–diesel W/D blends; in this case B30, to examine the readiness of RTES implementation should B30 rollout be carried out nationwide according to RMK-12 by 2025.

Experimental details

Test fuels

Base fuel used in this study is Malaysian Euro 2M low grade diesel fuel (D2M). It is commercially available in domestically in Malaysia and contain 10 vol% palm oil Fatty Acid Methyl Ester (POME) off the shelf. Therefore, since this study aims to investigate the effects of B30 biodiesel–diesel W/D blends, another 20 vol% of POME was added to D2M to form B30. The specifications of D2M and POME are indicated in Table 2. During preparation, 20 vol% of POME was measured and added to D2M before it is mixed using high shear mixer at a constant speed of 500 rpm in a closed container, to ensure mixture homogeneity. B30 was then immediately fed to the engine where it was let to run until the fuel lines were filled with B30 before any test run was conducted.

Table 2 Physicochemical properties of Malaysian Euro 2M low grade diesel (D2M) and palm oil based (POME).
Full size table

Meanwhile, to produce W/D emulsions of B30 (B30E), domestic tap water was used as the dispersed phase of the emulsion. The properties of tap water are explained in Table 3. Water percentages considered in this study was 10% (B30E10), 15% (B30E15), and 20% (B30E20). Higher water content was not desirable as higher than 20% will result in excessive vibration and engine stall. Meanwhile, physicochemical properties of B30 were not tested due to the need to continuously mix B30 to prevent coalescence and separation between D2M and POME.

Table 3 Physicochemical properties of Malaysian domestic tap water 27.
Full size table

W/D emulsions were produced by RTES using the updated RTES design developed by RTES Technology Sdn. Bhd. by removing ultrasonic agitator. Current RTES design incorporates a turbulence inducing mixing conduit and booster pumps. Figure 1 illustrates the updated RTES design. Detailed design was not revealed by the technology developer, however further information related to RTES can be acquired from their website25.

Figure 1
figure 1

Updated RTES design.

Full size image

Engine testing

Figure 2 depicts the engine testing setup for this study. Tests were done on a 5.9 L, 6-cylinder, turbocharged induction diesel engine with common rail injection. The engine is connected to a 100 kVA 4-pole 3-phase AC electric generator maintained at a constant speed of 1500 rpm to produce 420 V. The diesel-electric generator specifications are depicted in Table 4. Tests were conducted under variable electrical loads condition ranging from low load (5 kW), medium load (34 kW) and high load (64 kW) provided by an electrical load bank. The diesel-electric generator and load bank used in this study are shown in Fig. 3. The load bank is resistive-type, and the loading verification was done in-house according to Eq. (1)37 where power factor is assumed as 1 due to resistive loads. Thus, the results of the loading verification are explained in Table 5.

Figure 2
figure 2

Schematics of engine testing setup using RTES.

Full size image
Table 4 Specifications of diesel-electric generator.
Full size table
Figure 3
figure 3

Test engine and load bank configuration.

Full size image
Table 5 Electrical loading verification.
Full size table
$$Active power=\sqrt{3}\times Linevoltage \left(V\right)\times Linecurrent \left(A\right)\times power factor$$
(1)

As indicated by the manufacturer data in Table 4, peak power of the diesel-electric generator was achieved at 86 kW, therefore it is considered as 100% load. In relation to loading verification explained previously in Table 5, at 100% loading, an error between 9.08 and 9.92% was found. Meanwhile, at 75% (64 kW) loading and below, a maximum error of 6.41% was evaluated. Hence, for the purposes of this study, tests were conducted to a maximum load of 64 kW to ensure a stable power output with minimal error.

For engine performance and emissions testing, the engine is first warmed up using B30 until lubricant temperature stabilizes at approximately 60 °C. Following that, gate valve 1 is closed and gate valve 2 as shown in Fig. 2 is opened simultaneously, and the weight of main fuel tank is measured using electronic balance (accuracy ± 0.001 kg). This functions to bypass the main fuel tank by redirecting fuel flow to another fuel source. Upon completion of weighing, gate valve 1 is re-opened and gate valve 2 is re-closed and at the same time RTES system is activated to allow emulsification of B30 and water. Resultant B30E emulsion fuels are fed into the engine and tested for 6 min cycles under each load condition. This procedure is repeated for several cycles. Fuel consumed is measured by calculating the difference in weight prior to RTES activation and after each test cycles concluded. In each test cycle, NOx and CO exhaust gas emissions are measured using ECOM J2KN PRO gas analyzer while exhaust smoke opacity is measured using HORIBA MEXA-600S opacimeter. Both measuring equipment are depicted in Fig. 4. Technical specifications of both the gas analyzer and opacimeter are explained in Tables 6 and 7 respectively.

Figure 4
figure 4

Exhaust gas analysis measuring equipment.

Full size image
Table 6 Specifications of J2KN PRO gas analyzer.
Full size table
Table 7 Specifications of HORIBA MEXA-600S opacimeter.
Full size table

Experimental and uncertainty calculations

In this study, BTE and BSFC were calculated from fuel consumption data. Equation (2) 38 was used to evaluate BSFC:

$$BSFC=\frac{{\dot{m}}_{fuel}}{{P}_{b}}$$
(2)

where ṁfuel is the fuel mass flow rate measured in (g/h) while Pb is the engine brake power in kW. On the other hand, Eq. (3)38 was used to calculate BTE.

$$BTE,{\eta }_{thermal}=\frac{3600}{BSFC\times CV}\times 100$$
(3)

where BSFC is obtained from Eq. (2) measured in (g/kWh), and CV for D2M and POME verified in-house using CAL2K ECO bomb calorimeter. Meanwhile, CV for B30 was calculated by adding the CV of POME to D2M to achieve 30:70 blend ratio for biodiesel–diesel (42.9 MJ/kg). Furthermore, to calculate the CV for B30E10, B30E15, and B30E20, weighted average method was used, as explained in Eq. (4) 39.

$${CV}_{B30E}=\frac{\left({CV}_{B30}\times {mass}_{B30}\right)+\left({CV}_{water}\times {mass}_{water}\right)}{{mass}_{B30E}}$$
(4)

where CVwater is zero.

Furthermore, the calculation for uncertainty analysis used in this study is as shown in Eq. (5) 40.

$${\omega }_{R}=\sqrt{{\left(\frac{\delta R}{\delta {x}_{1}}{\omega }_{1}\right)}^{2}+{\left(\frac{\delta R}{\delta {x}_{2}}{\omega }_{2}\right)}^{2}+{\left(\frac{\delta R}{\delta {x}_{3}}{\omega }_{3}\right)}^{2}+\dots +{\left(\frac{\delta R}{\delta {x}_{n}}{\omega }_{n}\right)}^{2}}$$
(5)

where ωR is the total uncertainty of the experimental data, while ω1, ω2, ω3, to ωn represent independent variables. This equation is used to calculate the uncertainty of BTE and BSFC which consisted of independent variables such as ṁfuel, Pb, and fuel CV as evident in Eq. (3). For instance, uncertainty of BSFC is given by Eq. (6) 41.

$${\omega }_{BSFC}=\sqrt{{\left(\frac{\delta BSFC}{\delta {\dot{m}}_{fuel}}{\omega }_{{\dot{m}}_{fuel}}\right)}^{2}+{\left(\frac{\delta BSFC}{\delta {P}_{b}}{\omega }_{{P}_{b}}\right)}^{2}}$$
(6)

Therefore, the overall uncertainty of the experiments is as explained in Eq. (7)

$${\omega }_{Overall}=\sqrt{{uncertainty\, of\left(BSFC\right)}^{2}+{\left(BTE\right)}^{2}{+\left(CO\right)}^{2}{+\left(N{O}_{x}\right)}^{2}{+\left(Smoke\right)}^{2}} =\sqrt{{uncertainty\, of\left(1.38\right)}^{2}+{(1.38)}^{2}{+(2.28)}^{2}{+(3.87)}^{2}{+(2.63)}^{2}} = \pm 5.56\mathrm{\%}$$
(7)

Results and discussion

Brake thermal efficiency (BTE)

Figure 5 shows the effects of W/D emulsions towards engine BTE when operated under variable loads. It is obvious that BTE increases as engine load increases for all test fuels. At high engine load, D2M achieved 27.5% BTE while B30E10, B30E15 and B30E20 show significantly higher BTE at 32.8%, 35.6% and 37.4% respectively. B30E20 achieved the highest efficiency overall with a sizeable increase of 36.0% when compared to D2M at high engine load. Furthermore, at all tested engine loads, it was found that as water percentage in B30E emulsions increases, BTE increased as well. There are four postulations for this major improvement. Firstly, it may be due to micro-explosion effect that occurs when water droplets inside the emulsion fuel evaporated and tear the fuel droplet apart which can further improve combustion and thus producing higher in-cylinder pressure42,43. Although common rail fuel injection can produce very fine fuel droplets during pre-combustion, secondary atomization from the evaporation of water inside the B30E emulsion fuels still can occur and contribute towards enhancing air–fuel mixture. Secondly, emulsion fuels introduced into diesel engines tend to prolong ignition delay, which increases time for air–fuel mixing and evaporation, therefore improving combustion quality17,44,45. Thirdly, it may be due to the presence of oxygen in B30 biodiesel. Even though B30 possess lower calorific value when compared to D2M, which leads to increased fuel consumption as evident from previous documented research10,11,12, it does not mean that the combustion process itself was not efficient. The presence of oxygen in biodiesels assist combustion at fuel rich regions, therefore improving combustion efficiency. However, in most cases the improvement of combustion efficiency could not offset the deficiency of energy content within the biodiesel fuel itself and hence resulted in higher BSFC. Finally, it is possible that evaporation and expansion of water droplets within the emulsion fuel during combustion process increased the overall rate of heat release when compared to conventional diesel. Water do not carry any energy for the purpose of combustion, and act only as an expansion agent within the combustion cylinder by absorbing the heat released during combustion. As the content of water (as the expansion agent) increases, the higher is the rate of heat release inside the combustion cylinder. In fact, in-cylinder pressure could be higher as compared to conventional diesel if the start of ignition can be modified to be at the same timing30,46. In contrast, if there is too much water in the emulsion fuel, the absorption of heat by water particles will start to quench some part of the chemical reactions occurring during combustion, therefore limiting the maximum water content that could be introduced into the emulsion fuel. In this study, it was found that 20 wt% of water was the optimum water percentage within the W/D emulsion that can act as the expansion agent without significantly affecting chemical reaction of combustion process.

Figure 5
figure 5

BTE of test fuels at different engine loads.

Full size image

Brake specific fuel consumption (BSFC)

In this study, for the purpose of calculating BSFC, only B30 is considered as the fuel. This is because water is not a combustible substance. This method was chosen with reference to previous studies on BSFC behavior of W/D emulsion fuels28,32,47. Therefore, in accordance with Eq. (2), BSFC of D2M and various water percentage B30E when subjected to increasing engine loads are as illustrated in Fig. 6. It is observed that BSFC decreases as the engine load increases. This signifies that engine combustion efficiency is enhanced at higher engine loads17. Furthermore, it is also observed that at each tested engine load, BSFC improved as water percentage in B30E increases. Most notably at low engine load, where BSFC for B30E10, B30E15 and B30E20 fuels showed significant decrease by 3.67%, 7.43% and 8.70% respectively, as compared to D2M. Meanwhile, at high engine load of 64 kW, B30E15 displayed maximum BSFC reduction with a 7.19% improvement. Considering all engine loads, average reduction of each B30E emulsion fuels is 1.18%, 4.63% and 3.6% for B30E10, B30E15 and B30E20 respectively. This observed reduction of BSFC can be considered as significant, since a previous report observed a marginal BSFC increase when B30 biodiesel was fueled in a common rail injection diesel vehicle as opposed to B1010. Despite having a calorific value deficit of about 16.8% when compared to D2M, significantly lower BSFC and higher BTE achieved with B30E15 showed that micro explosion effect in W/D emulsion is sufficient to overcome the same engine loads with lower amount of fuel.

Figure 6
figure 6

BSFC of test fuels at different engine loads.

Full size image

NOx emissions

NOx emissions of D2M and B30E under increasing engine loads are as shown in Fig. 7. It is obvious that NOx emission increases as engine load increases. At the low load engine condition of 5 kW, the average NOx emission is at the minimum level of 60.4 g/kWh. As engine load increases, average NOx emissions increase by measures of approximately 400% and 950% for medium and high loads respectively. NOx gases are formed typically at high temperatures exceeding 1800 K through Zeldovich mechanism48. Therefore, as engine load increases, combustion cylinder temperature increases, resulting in a higher rate of NOx formation. Furthermore, as water content in B30E increases, NOx emissions decrease at all tested engine loads. Particularly at high load, D2M emitted 672 g/kWh of NOx while B30E10, B30E15 and B30E20 emulsions produced lower NOx by margins of 5.95%, 5.65% and 11.6% respectively. Water droplets present in B30E fuels resulted in lower flame temperatures during combustion as the result of latent heat absorption by water particles. Hence, NOx formation by Zeldovich mechanism was restricted49. In addition, as water content in B30E increases, the amount of B30 fuel injected per volume is considerably lower, reducing the amount of combustion by-products, and further limiting NOx formation42. Moreover, as water percentage increases, chances of more oxygen molecules ionize to form of hydroxyl (OH) radicals increases significantly, leading to lower NOx formation29. Nonetheless, reduction of NOx observed in this study is rather minimum when compared to previous studies17,22,34. This observation could be explained for two reasons. Firstly, B30E emulsions contain higher biodiesel content which translates to higher oxygen content in the fuel. Therefore, even though it is expected that B30 should produce higher NOx emissions (due to oxygen presence), since water droplets are present in B30E, formation of NOx was suppressed. Secondly, it is possible that due to higher combustion temperature promoted by common rail fuel injection coupled with a turbocharged induction air intake system, micro explosion of minute water particles in B30E emulsions could not effectively reduce combustion temperature in a magnitude observed in previous studies. These assertions were substantiated in a separate study which documented that a common rail turbocharged engine fueled with polyoxymethylene dimethyl ethers-diesel, a highly oxygenated fuel, was unable to effectively suppress NOx formation under various engine loads50.

Figure 7
figure 7

NOx emission of test fuels at different engine loads.

Full size image

CO emissions

Figure 8 shows the CO emissions of D2M and B30E under various loads. It is evident from the figure that CO emissions for all fuels are at the highest at low engine load. This is due to lower combustion cylinder temperature as the result of increase in heat loss per cycle51. Therefore, the rate of oxidation from CO to CO2 below 1400 K decreases17. Meanwhile, a slight increase is observed at high engine load, where CO emissions for all test fuels are slightly higher as compared to medium load. It can be explained that, at higher engine load, excess fuel is injected into the combustion cylinder which lead to stratified rich mixture regions. This in turn results in lesser contact between fuel and oxygen, leading to poor combustion within these regions. This notion was corroborated by a previous study that reported CO emissions increase in higher engine loads due to oxygen deficiency at the end of fuel jet impinged on the cylinder wall52. Nonetheless, it is obvious that at the low engine load, B30E emulsions displayed substantially higher CO emissions as compared to D2M. This is due to cooling effect which occurs during combustion of W/D emulsion, which encouraged incomplete oxidation of CO to CO2 in the presence of water droplets51,53. However, at high engine load B30E emulsions showcased much lower CO emissions than D2M. Higher engine loads are characterized by higher in-cylinder temperatures and pressures, where it can be argued that micro explosion occurs more violently, resulting in finer fuel droplets distribution, ultimately improving combustion17.

Figure 8
figure 8

CO emission of test fuels at different engine loads.

Full size image

Smoke opacity

Figure 9 illustrates smoke opacity profile of D2M and various B30E under increasing engine load. It is clear that smoke opacity increases as engine load increases. This is because, more fuel is injected into the engine in high engine loads, causing air–fuel ratio to decrease, hence resulting in formation of smoke due to incomplete combustion54. Generally, it can be seen at high engine load, smoke opacity becomes lower as water content in B30 increases. In fact, maximum smoke opacity reduction is achieved by B30E15 under high loads, with a magnitude of 61.5% reduction with respect to D2M. Again, it is obvious that the role of micro explosion is very effective in improving fuel atomization, achieving higher combustion quality22 in high engine load environment characterized by higher in-cylinder pressure and temperature. Another contributing factor is due to the presence of higher OH radicals in B30E, air entrainment is enriched, further reducing smoke formation55. Similar trends were observed previously using D2M derived W/D emulsions where a maximum smoke opacity reduction of 87.0% was achieved by 21.8 wt% water content when fueled into the same test engine32. In addition, this trend is also observed for other types of biodiesel–diesel emulsion such as Nerium biodiesel–diesel, which showed up to 12.96% reduction in smoke opacity56, attributed by the effects of micro explosion. It can also be noted that the presence of oxygen in POME (hence B30) contributed towards more efficient combustion and suppressed formation of smoke57 In short, RTES implementation has been effective in reducing smoke emissions in common rail injection engines, which has negated suspicions that the effect of micro explosion is less pronounced in high pressure fuel spray of common rail injection. Furthermore, similar to CO emissions trend presented in Fig. 8, higher smoke emissions are detected in low engine load when fueled with B30E where the effect of micro explosion is less significant.

Figure 9
figure 9

Smoke opacity of test fuels at different engine loads.

Full size image

Conclusion

This research attempted to improve performance and emissions of a 100 kVA common rail fuel injection, turbocharged diesel electric generator by using water-in-diesel emulsion fuel derived from higher biodiesel–diesel blends (B30) without addition of any emulsifier or surfactant. The base fuel chosen was a low-grade Malaysian Euro 2 diesel fuel (D2M) and three emulsified fuels of B30 produced by mixing D2M with POME (B30E). Water percentages considered were 10 wt% (B30E10), 15 wt% (B30E15) and 20 wt% (B30E20). B30E were produced using Real-Time Non-Surfactant Emulsion Fuel Supply System (RTES) immediately before entering the common rail fuel injection system.

The following conclusions are drawn from the experiment:

  1. 1.

    Brake Thermal Efficiency (BTE) at each tested engine load increased as water percentage in B30E emulsion fuels increased. The highest increment was observed at high engine load by B30E20 with 36.0% increment as compared to D2M.

  2. 2.

    Generally, BSFC reduced for all B30E emulsion fuels except for a slight increase at medium engine load, as compared to D2M. The maximum reduction was observed at low engine load with 8.70% reduction by B30E20. At high engine load, B30E15 showed a maximum reduction of 7.19%.

  3. 3.

    As water percentage in B30E emulsion fuels increased, NOx emissions were reduced. However, the margin the reductions observed were not as pronounced as the ones observed in previous studies, possibly due to higher combustion temperatures when using forced induction common rail injection systems.

  4. 4.

    B30E emulsion fuels shows higher CO emission concentration at low engine load and lower CO emission concentration at high engine load than D2M. Combustion temperature at different engine loads influence the oxidation of CO to CO2 and strength of micro-explosion of B30E emulsion fuels.

  5. 5.

    Smoke opacity of B30E emulsion fuels reduced significantly as compared to D2M at high engine load. This proves that micro explosion restricted formation of soot particles by improving combustion through secondary fuel atomization. However, the effect of micro-explosion was weaker at low and medium engine loads.

In short, higher blends biodiesel–diesel W/D emulsions produced by RTES was able to improve engine thermal efficiency and fuel consumption as well as reducing NOx and soot emissions when installed in a modern diesel engine. Based on the results of this study, RTES implementation is recommended when B30 biodiesel mandate is enforced in Malaysia in the near future by 2025.