Experimental Study of Port Water Injection on GDI Engine Fuel Economics and Emissions

Abstract

This experimental study investigates the impact of water injection into intake ports on combustion and emissions in gasoline engines. It also examines particle size distribution at various water-to-fuel ratios and explores the combined effects of water injection and compression ratios in gasoline engines. The results indicate that water injection effectively mitigates engine knock, reduces peak firing pressures, and moderates heat release rates through charge cooling. Advancing ignition timing with water injection advances combustion, resulting in reduced specific fuel consumption, particularly under moderate load conditions. Water injection lowers NOx emissions by reducing combustion temperatures but increases unburned THC emissions due to inhibited oxidation reaction rates. Minor effects were observed on CO emissions. Furthermore, particle numbers were significantly reduced with water injection, particularly in the nucleation mode particles. The simultaneous application of a higher compression ratio and water injection yields substantial improvements in fuel consumption with minimal impact on NOx and THC emissions.

1. Introduction

Climate change driven by greenhouse gas emissions poses one of the most serious threats facing humanity. Based on the current situation, the average global temperature is projected to rise by approximately 2 °C over the next 80 years relative to preindustrial levels. This will likely cause substantial adverse impacts including sea level rise, more extreme weather events, ecosystem disruptions, and increased spread of infectious diseases.¹ Currently, the primary sources of CO₂ emissions include the industrial sector, encompassing power generation, iron and steel production, cement manufacturing, among others, and the transportation sector, predominantly characterized by internal combustion engine vehicles.² These vehicles predominantly rely on internal combustion engines burning fossil fuels like gasoline and diesel, which release CO₂ as a byproduct of combustion. Reducing CO₂ emissions from transportation is vital to mitigate climate change risks. This requires aggressive electrification of vehicles, improved efficiency of internal combustion engines, and transitioning to lower carbon neutral fuels. For internal combustion engines, strategies like engine downsizing,³ combustion optimization,⁴ and electrification⁵ can reduce fuel consumption and CO₂ emissions from conventional fossil fuel-powered vehicles in the near term.

In spark-ignition gasoline engines, there are two primary approaches to improve thermal efficiency—increasing the compression ratio and diluting the intake charge.⁶ Higher compression ratios promote an increased thermal efficiency by allowing more expansion work to be extracted during the power stroke. However, excessive compression ratios can lead to undesired engine knock due to preignition.⁷ Knocking can be mitigated by retarding spark timing in order to reduce the in-cylinder temperature. However, this will inevitably lead to a decrease in the thermal efficiency. Diluting the intake charge via excess air,⁸ residual gases,⁹ or other lower boiling point liquid¹⁰ reduces the propensity for knock by lowering the charge temperature during compression. This allows spark timing to be advanced closer to the top dead center (TDC) while avoiding knock, enabling greater efficiency. Dilution increases the specific heat capacity of the intake charge, which requires greater energy input to achieve the same temperature rise during compression, compensating for the reduced chemical energy and further suppressing the knock propensity.

Water injection in gasoline engines can provide significant charge cooling and knock mitigation benefits.^11,12 Water has a higher latent heat of vaporization compared with common engine fuels and diluents. As liquid water vaporizes upon injection, substantial heat is absorbed from the air charge, markedly reducing the temperature during compression. This increases the preignition resistance and prevents knock, allowing for more advanced, efficient combustion phasing. Additionally, the water vapor acts as a diluent during combustion similar to that for cooled exhaust gas recirculation (EGR). By absorbing energy during vaporization and increasing the specific heat capacity of the charge, water vapor reduces the peak combustion temperatures. This suppresses the formation of oxides of nitrogen (NOx) emissions which are thermally sensitive.¹³ The reduced temperatures and dilution effect also inhibit abnormal combustion and knock in a manner similar to EGR.

For water injection implementation, two primary methods were used, water direct injection and port water injection. With water direct injection, liquid water is sprayed directly into the combustion chamber via an additional high-pressure injection system. In contrast, port water injection injects water into the intake air stream upstream of the intake valves. Prior work by Kim et al.¹⁴ demonstrated that direct in-cylinder water injection could reduce combustion temperatures and mitigate knock, enabling increased output torque. Valero-Marco et al.¹⁵ demonstrated the potential of water direct injection to increase the maximum load capability of compression ignition engines. Results showed that direct in-cylinder water injection enabled suppression of the knock tendencies despite higher compression, allowing extension of the load limit. Numerical modeling by Falfari et al.¹⁶ using computational fluid dynamics provided insights into the effects of water injection on in-cylinder thermal conditions and combustion characteristics. Water injection substantially reduced peak gas temperatures during combustion due to the vaporization cooling effect, which slowed the flame front propagation speed as reaction rates decreased. Subramanian¹⁷ found that under high load conditions, both direct water injection and port water injection techniques could substantially reduce NO emissions to a similar extent.

Due to the limited space on the cylinder head, the DI water injector may occupy the room of the spark plug or DI gasoline injector, which have to make a compromise between higher specific power and larger valve diameter. However, port water injection can make full use of existing gasoline port fuel injection system, which making it the most feasible approach for near-term production implementation.¹⁸ According to the results of the investigation by Iacobacci et al.,¹⁹ water injection enabled substantial advancements in spark timing by mitigating knock tendencies. Compared to baseline tests without water injection, port water injection reduced fuel consumption by 6–12% across the speed range based on brake-specific fuel consumption (BSFC) measurements. Furthermore, Hoppe et al.²⁰ showed through experiments on a single-cylinder engine that port water injection could increase thermal efficiency by up to 16% at 2000 rpm and 2.26 MPa IMEP by mitigating knock and enabling advanced, efficient combustion phasing. Besides, in spark ignition engines, laminar burning velocity is a fundamental parameter characterizing oxidizer-fuel mixture reactivity and diffusion combustion. Numerical simulations by Mazas et al.²¹ examined the effects of water vapor addition on the characteristics of laminar coflow diffusion flames. Their modeling results showed that injecting water vapor into the air stream had minor direct chemical kinetic effects on the laminar burning velocity. However, some researchers reported that water injection lengthened the CA90, and the combustion duration was lengthened with the load increase.^14,22,23

Meeting increasingly stringent emission regulations to mitigate environmental impacts represents a pivotal challenge for internal combustion engines. Tailpipe pollutant emissions, such as nitrogen oxides (NOx), particulate matter, carbon monoxide (CO), and unburned hydrocarbons (UHCs), must be minimized to reduce air pollution and comply with emissions standards. Water addition has been investigated for emission control in combustion systems.^24,25 Numerous studies have demonstrated that injecting water into the combustion chamber or upstream of the flame zone can effectively reduce nitric oxide (NOx) emissions.^14,26 However, this NOx reduction is often accompanied by increases in unburned hydrocarbon (HC) emissions.^11,27

This study aims to computationally investigate the impact of port water injection on emissions and performance in gasoline engines. Additionally, the study examines the influence of the water-to-fuel ratio on particle size distribution and adjusts the ignition timing based on the knock boundary, which is rarely reported in previous studies. It is worth noting that a high compression ratio increases the density of the fuel-air mixture and the turbulence in the combustion chamber, leading to higher cylinder pressure and faster combustion rates. Comparing the fuel consumption and emissions among various water-to-fuel ratios under different operating conditions, the combined effects of water injection and compression ratio on gasoline engine are explored.

2. Experimental Section

2.1. Experimental Engine

In this study, a four-cylinder Miller 2.0 L gasoline engine equipped with a 35 MPa high-pressure fuel injection system was applied. This engine is equipped with one side-mounted DI injector per cylinder and is modified for water port injection. The port-fuel injectors are mounted on the intake manifold, positioned upstream of the intake port, and the water spray is targeted at the back of the intake valve. The control parameters of water injection, such as the start of injection and injection width, could be synchronously adjusted using INCA software, while the water injection pressure during the experiments remained constant at 0.8 MPa. In order to explore the influence of water injection technology on engine performance under different compression ratios, the engine’s compression ratio was adjusted by means of piston replacement. Relatively, higher compression ratios corresponded to greater piston protrusion, and conversely, lower compression ratios led to lesser protrusion. The key technical parameters of the experimental engine are presented in Table 1.

Table 1. Engine Specifications.

displacement volume [L]	2.0
bore/stroke [mm/mm]	75/85
compression ratio	11, 12.1, 13.3
number of cylinders	4
injection system	side mounted DI injector
rated power/speed [kW/rpm]	150/5000

2.2. Test-Cell Layout

The dynamometer utilized in the test is an AVL electric dynamometer, while the AVL 735S is employed as the fuel consumption meter. The fuel consumption meter is equipped with a 753C temperature control device for regulating fuel temperature. To control the intake temperature, an intake system with a water-cooled intercooler was applied, which maintains the intake temperature by flexibly controlling the coolant flow rate and temperature. Injection timing, fuel injection quantity, and ignition timing were adjusted by the INCA software. For cycle-based in-cylinder pressure recording, a piezoelectric pressure sensor (Kistler 6115c) is integrated within the combustion chamber. In-cylinder pressure data acquisition and analysis were performed using the AVL INDICOM combustion analyzer. An average of the pressure data for 200 cycles was used for indicator analysis. The engine exhaust emission sampling line was installed before the three-way catalyst. Initial engine emissions (CO, HC, NOx, etc.) are measured through the HORIBA MEXA-7400 DEGR system. The engine’s excess air ratio (lambda) is calculated and monitored by an ETAS ES630 system with a BOSCH LSU 4.9 exhaust oxygen content sensor. To investigate the impact of water injection on engine performance and combustion processes, the water-to-fuel ratio (the ratio of the mass of water injected into the intake to the mass of fuel injected into the cylinder) is defined as the key parameter, abbreviated as W/F. For precise water measurement, the ONO SOKKI DF210 volumetric flowmeter is employed to accurately measure the quantity of injected water. Figure 1 demonstrates the schematic of the test rig and experimental apparatus. The main specifications of the test equipment used in the tests are presented in Table 2.

Figure 1.

Schematic of the test rig configuration.

Table 2. Engine Specifications.

test equipment	model	manufacturer	uncertainties
dynamometer	S22–2	AVL	torque: ±1 N m
			speed: ±2 rpm
fuel-flow meter	DF210	ONO-SOKKI	±0.2%
pressure transducer	6115C	KISTLER	±1%
exhaust gas analyzer	MEXA-7400DEGR	HORIBA	±0.5% FS injector
lambda meter	ES630	ETAS	±0.05(<1.6)

2.3. Engine Test Conditions

Experiments were conducted at fixed engine speeds of 2500 rpm, brake mean effective pressures (BMEPs) of 0.5, 1.1, and 1.7 MPa, and wide open throttle (WOT) conditions for this study. Gasoline with an octane number of 92 was used for all tests. The engine coolant and lubricating oil temperatures were maintained at 80 °C. The intake air temperature was fixed at 30 °C using a charge intercooler. The equivalence ratio φ, measured by an exhaust lambda sensor, was kept stoichiometric at 1.0. Prior to data acquisition, the engine was warmed up for at least 30 min, ascertained by coolant and oil temperature sensors. External control of the water injection system was achieved by using a driver that facilitated the management of water injection timing and intervals. Water, maintained at a temperature of 25 °C, was injected into the intake stroke, and an exploration of water-to-gasoline ratios was conducted, ranging from 0.10 to 0.30 mass fractions, as depicted in Figure 2. The impact of water injection on mitigating knock was also investigated under engine speeds of 2500 and 5000 rpm, advancing the spark timing while maintaining a constant Lambda value until the onset of knock.

Figure 2.

Schematic for water injection timing calculations.

2.4. Test Methodology

To ensure the repeatability of the experimental results, in-cylinder pressure data were collected over 200 consecutive engine cycles for each operating condition. The pressure data were analyzed by using combustion analysis software to obtain key combustion parameters for comparing engine combustion characteristics for different cases. In this study, CA50 is defined as the crank angle corresponding to 50% cumulative heat release representing combustion phasing. CA10 and CA90 denote 10% and 90% cumulative heat release points, respectively. The duration from CA10 to CA90, denoted as CA10–90, represents the combustion duration. The BSFC is used to evaluate fuel economies, while the coefficient of variation of indicated mean effective pressure (COV_IMEP) given by eq 1 is used to assess combustion stability.

1

where represents the average IMEP value over 200 cycles and σ_IMEP represents the standard deviation in the same cycles. A COV of IMEP above 3% indicates unstable combustion conditions. Besides CA50, CA10–90, BSFC, COV_IMEP, and other important parameters, such as peak firing pressure, maximum pressure rise rate (MPRR), and combustion noise, can also be extracted from the in-cylinder pressure data to further analyze the impact of water injection strategies on engine combustion characteristics.

The maximum amplitude of pressure oscillations (MAPO) is used as a metric to assess the knock intensity. The in-cylinder pressure data is first preprocessed using a 5–20 kHz band-pass filter to extract the high-frequency pressure oscillation signals associated with the knock from the raw pressure trace, as shown in Figure 3, which can be illustrated by.²⁸ The MAPO value is then calculated using eq 2

2

where p~ is the band-pass filtered cylinder pressure. The MAPO represents the peak amplitude of the filtered pressure oscillations, and higher values indicate a more intense knock.

Figure 3.

Definition of MAPO from the cylinder pressure curves.

3. Results and Discussion

As mentioned above, the aim of this work was to study the effect of port water injection on the combustion and emissions of a gasoline engine under various conditions. Particle number emissions can also be compared with various water injection fractions.

3.1. Effect of Port Water Injection on Combustion

In direct injection gasoline engines, the native properties of gasoline fuel may promote abnormal combustion under higher load operating conditions. With higher intake pressures and combustion temperatures, mixtures are easy to autoignition and consequently rapid combustion, which leads to undesirable engine knock. Knock tends to occur more readily with gasoline relative to other fuels, such as diesel, due to the presence of more octane-sensitive paraffinic components. This autoignition creates pressure waves, which may lead to potential engine damage if severe. Therefore, knowing how to control the cylinder temperature is the key to inhibiting the knock. In this paper, experiments were conducted at speeds of 2500 and 5000 rpm in order to clarify the effect of port water injection on engine knock combustion.

As can be seen in Figure 4, under the load of BMEP 1.7 MPa, a higher fraction of water reduces the peak firing pressure, the combustion process is obviously retarded with more water injected into the cylinder, and knocking has been improved significantly, while the spark timing is the same. Second, as presented in Figure 5, the heat release rate curve also tends to be smoother with more water injected. A similar trend was reported by Hoth et al.²⁹ It is related to the fact that the water injected into the intake port vaporized rapidly in the high-temperature environment during the compression stroke as fresh air was drawn into the cylinder. In this process, the phase change of liquid water to vapor absorbs a large amount of heat, reducing the in-cylinder charge temperature. Consequently, the initial combustion temperature is lowered at the end of the compression stroke. Hence, the overall temperature increase rate during combustion becomes slower as the bulk gas mixture is cooler. This helps suppress knock-down tendencies to some extent. Furthermore, a decreased temperature leads to a slower combustion speed. With fixed spark timing, this manifests as an overall retardation of the combustion event, as observed from the cylinder pressure curves with a lower peak pressure. The reduced burn rates at lower temperatures are attributed to a slower chemical reaction process.

Figure 4.

In-cylinder pressure traces for different water-to-fuel ratios.

Figure 5.

Heat release rate traces for different water-to-fuel ratios.

Knock characteristics during combustion are evaluated by a statistical analysis of high-frequency pressure fluctuations. The MPRR and the maximum amplitude of pressure are used as indicators to evaluate the knock characteristics. The MAPO metric is calculated from consecutive engine cycles to quantify knock intensity. Typically, a MAPO value exceeding 1 bar is considered the knock limit for gasoline engines above which significant knockout combustion occurs. The MAPO and MPRR values under different water-to-fuel ratio conditions are shown in Figure 6. Due to the presence of knocking during the combustion process, the MPRR is higher when water is not injected. Without water injection, MAPO frequently exceeds 1 bar across 200 consecutive test cycles, with some cycles approaching around 2.5 bar. However, with increasing water mass fractions, MAPO decreases markedly. Above 10% water-to-fuel ratio, MAPO remains below 0.5 bar, with no evident knock observed in continuous testing. This demonstrates that water injection could make use of the charge cooling effect of liquid water vaporization to reduce combustion temperatures, thereby positively mitigating knock tendencies.

Figure 6.

MAPO and MPRR for consecutive test cycles with various water-to-fuel ratios.

To further investigate the potential of intake port water injection to improve the combustion process, the spark timing was adjusted under various water-to-fuel ratio conditions to achieve the same knock limit for all cases. The resulting differences in combustion characteristics were compared. Figures 7 and 8 show the correlation between BMEP and spark timing under various water-to-fuel ratios. With fixed spark timing, increasing the water-to-fuel ratio reduces the combustion temperature and slows down the combustion rate, retarding the overall combustion process. The ignition delay, defined as the duration from spark timing to CA10, extends with an increase in the water fuel ratio. This is primarily due to the endothermic vaporization of water, which lowers the in-cylinder temperature and further slows down the rate of chemical reactions during the initial phase of combustion, as can be seen in Figure 9. This results in a markedly reduced BMEP, especially at higher water levels. Advancing spark timing after water injection advances the whole combustion process with an earlier CA50. Hence, the peak firing pressure increases and results in higher work output, as shown in Figures 10 and 11. At 2500 rpm, BMEP = 1.7 MPa, the required spark advance was around 3 °CA at a water-to-fuel ratio of 0.3 compared to no water injection. Additionally, the figures show that the BMEP increase due to advanced timing diminishes with a rising water content. This suggests that, with a higher water fraction, the vaporized water dilutes the charge by acting as a chemically inert gas, reducing reactant concentrations and slowing reaction rates to some extent. The prolonged ignition delay necessitates further spark advance to set aside adequate time for the main heat release phase to achieve higher loads.

Figure 7.

Effects of water injection on the BMEP with different spark timings.

Figure 8.

Effects of water injection on the peak pressure with different spark timings.

Figure 9.

Effects of water injection on the ignition delay (CA10-spark timing) with different spark timings.

Figure 10.

Effects of water injection on CA50 with different spark timings.

Figure 11.

Effects of water injection on the combustion duration with different spark timings.

Figure 10 also shows that advancing the spark timing under various water-to-fuel ratio conditions may shift the overall combustion process earlier. Moreover, due to the improvement in the constant volume degree, the combustion duration is shortened. However, when the water-to-fuel ratio reaches 0.3, further advancing the spark timing to −15 °CA after the top dead center (ATDC) results in an increase in the combustion duration. This may be due to the fact that although increasing the water-to-fuel ratio can lower the combustion temperature, mitigate knock, and enable earlier spark timing to bring the combustion process as close as possible to the TDC for improved constant volume degree and shortened combustion duration, an excessively high water-to-fuel ratio will decelerate the reaction rates during combustion. Moreover, under the combined effects of these two factors, when the water-to-fuel ratio reaches 0.3, further advancing the ignition timing to −15 °CA ATDC, the improvement effect on the combustion duration due to the constant volume degree gradually weakens. This phenomenon can be explained by the two competing effects of water injection on flame propagation. On one hand, water injection cools the charge temperature, which tends to slow down flame speed. On the other hand, it suppresses knock which allows advancing ignition timing closer to TDC which tends to increase flame speed. At lower water-to-fuel ratios, the positive effect of advanced ignition timing dominates, leading to faster overall flame propagation. However, at a very high-water content such as 0.3, the negative effect of reduced temperature becomes more significant, slowing down the flame speed and prolonging combustion duration despite the advanced timing.

Figure 12 illustrates the trends of BSFC and the Coefficient of Variation of Indicated Mean Effective Pressure (COVIMEP) under varying load conditions and water-to-fuel ratios. In this experiment, to ensure the reliability of the data, the spark timing for different water-to-fuel ratios was adjusted to maintain combustion at the knock limit (MAPO < 1 bar). It is evident that under lower loads, BSFC consistently increases as the water-to-fuel ratio rises. However, for medium and higher loads, BSFC decreases as the water-to-fuel ratio increases, up to approximately 0.2. Beyond this point, there is a slight increase in the BSFC at higher ratios. In the case of gasoline engines, the combustion temperature remains relatively low under lower loads. Water injection further reduces the end-of-compression temperature, which could lead to ignition challenges and reduced combustion stability. With a water-to-fuel ratio of 0.3, BSFC increased by approximately 13% compared to the condition without water injection.

Figure 12.

Comparison of BSFC and COV_IMEP under different load and water-to-fuel ratios: (a) BSFC; (b) COV_IMEP.

At medium and high loads, CA50 is typically retarded due to knock limits. The combustion event primarily occurs during the expansion stroke, while the piston is moving downward after TDC. This reduces the constant volume degree of combustion, as less heat release happens near the TDC, lowering the thermal efficiency. Knock avoidance requires retarding the spark timing to prevent autoignition of the mixture ahead of the flame. However, the resulting delayed phasing shifts the combustion process further from the TDC into the expansion stroke. The reduction in constant volume combustion decreases the thermal efficiency. The work output is also reducing as the pressure rise from the delayed burn occurs later in the cycle when the piston velocity is higher. With water injection, the improved knock resistance advances the overall combustion process and significantly enhances combustion stability, markedly reducing BSFC. At a ratio of 0.3, BSFC decreased by about 10% compared to that without water injection.

3.2. Effect of Port Water Injection on Emissions

To further elucidate the impact of port water injection on emissions from gasoline engines, experiments were conducted to analyze the influence of the water-to-fuel ratio on emissions under different load conditions. Figure 13 illustrates the NOx emissions at various water-to-fuel ratios. It is evident that NOx production varies across different water-to-fuel ratios. As the water-to-fuel ratio increases, the delayed combustion timing and, subsequently, lower combustion temperatures contribute to a relatively reduced generation of NOx.

Figure 13.

NOx emission with various water-to-fuel ratios under different loads.

Figure 14 compares the total hydrocarbon (THC) and CO emissions at various water-to-fuel ratios, along with the combustion efficiency for each condition, illustrating the trend in efficiency changes under these different ratios. It can be seen that THC emissions increase with rising water content, with more significant unburned HC emissions at higher ratios. The main reason for this phenomenon is that when large amounts of liquid water enter the cylinder and vaporize rapidly, absorbing heat and transitioning to the gaseous state, the vaporized water cannot participate in the combustion reactions. This reduces the probability of molecular collisions between reacting species during the combustion process, increasing the combustion incompleteness. Consequently, the combustion efficiency is decreased, elevating unburned THC emissions. The reduced combustion temperatures from vaporization limit the reaction rates and oxidation of the fuel. However, the stoichiometric air/fuel ratios provide adequate oxygen for eventual CO oxidation by the end of combustion, as evidenced by the minor impact on CO.

Figure 14.

Combustion efficiency and THC/CO emission with various water-to-fuel ratios under different loads.

The results related to the particle size distribution are presented in this section. A Cambustion DMS500 particle analyzer is applied in the test. The figure presented in Figure 15 illustrates the particle size distribution across a range of water-to-fuel ratios, spanning from 0 to 0.3, at two engine loads: 2500 rpm with 0.5 and 1.7 MPa. As can be seen in this figure, for the engine load of 0.5 MPa, the particle size distribution changed from the trend of monotonical decrease. When the engine load increased to 1.7 MPa, the particle size distribution followed a single peak distribution pattern for different water-to-fuel ratios. Advancing the spark timing and the addition of water into the intake port at the same time decrease the particle number, especially for the nuclear mode particles. The peak particle size increased from around 8 nm at a lower load to 13 nm at a larger load. The diminishing of the particles around 8 nm can be explained by the more sufficient combustion, which effectively reduced the soot formation, and consequently, the formation of nucleation mode particles would be suppressed.

Figure 15.

Particle size distribution with various water-to-fuel ratios under different loads: (a) 2500 rpm, 0.5 MPa; (b) 2500 rpm, 1.7 MPa.

3.3. Effect of Port Water Injection on Emissions

Improvements in engine performance can be obtained by several means, such as increasing the compression ratio. High compression ratio increases mixture density at the end of the compression stroke and charge temperature in the combustion chamber, resulting in a higher combustion rate. Figure 16 shows the differences in BSFC between 12.1:1, 13.3:1, and 11:1. Figure 16 demonstrates that for various water-to-fuel ratios, an increase in the compression ratio leads to higher specific fuel consumption. Specifically, a compression ratio of 11:1 results in lower fuel consumption compared with compression ratios of 12.1:1 and 13.3:1, with these differences becoming more prominent at higher loads and lower water-to-fuel ratios. However, different outcomes are observed under lighter loads and higher water-to-fuel ratios. The most significant improvement in fuel consumption is observed with a compression ratio of 13.3:1 under light and moderate loads, especially when a higher water-to-fuel ratio is used. This is because, as observed previously, the gasoline knock resistance limits the compression ratio. In comparison to compression ratios of 11:1 and 12.1:1, BSFC notably decreases with greater water injection under lower and moderate loads when a compression ratio of 13.3 is employed. Advanced spark timing and a higher water-to-fuel ratio may reduce the tendency for knock with a higher compression ratio. However, a limited water-to-fuel ratio cannot effectively reduce in-cylinder temperatures under heavier loads.

Figure 16.

Differences in BSFC between 12.1:1, 13.3:1, and 11:1: (a) delta BSFC of 12.1:1 to 11:1; (b) delta BSFC of 13.3:1 to 11:1.

The differences in NOx and THC emissions between with and without port water injection under various operating conditions for the three compression ratios are reported in Figures 17 and 18. Due to the slight influence of the water-to-fuel ratio on CO emissions, according to the last section, NOx and THC have been focused on the comparison with various water-to-fuel ratios for different compression ratios. As depicted in these figures, the effect of the water-to-fuel ratio on NOx emissions varies under different load and compression ratio conditions. At middle speed and load, combustion is affected by both advanced spark timing and charge cooling from water vaporization after port injection. With a compression ratio of 13.3, NOx emissions increase slightly compared to those without water injection when the water-to-fuel ratio is 0.1, unlike the trends for other compression ratios. However, for a higher speed and load, such as the WOT operating condition at 5000 rpm, knock limitations prevent operation at a compression ratio of 13.3 without water injection. Moreover, the reduced temperature and dilute effect of the water vapor mitigate knock tendencies, enabling significant spark advance and higher in-cylinder combustion temperatures, increasing NOx emissions to some extent. Furthermore, not fully vaporized liquid water may also be one of the potential causes due to the poor atomization level of liquid water at higher engine speeds, which has limited time for vaporization until the main combustion process proceeds. At high speeds and loads in Figure 18, the elevated combustion temperatures result in markedly decreased THC emissions.

Figure 17.

Differences in NOx emissions with and without port water injection under various operating conditions.

Figure 18.

Differences in THC emissions between with and without port water injection under various operating conditions.

4. Conclusions

The effects of intake port water injection on combustion and emissions in gasoline engines were explored experimentally to assess the potential of this technique for improving combustion and reducing emissions. The main conclusions are summarized as follows:

• (1)Water injection suppresses knock, allowing advanced spark timing to improve combustion phasing. However, excessive water addition can slow flame propagation.
• (2)Experiments found that water injection reduces peak firing pressures and heat release rates, mitigating knock at high loads. Advancing timing with water injection advances combustion phasing, increasing peak pressures and BMEP. However, benefits diminish at very high water content due to slowed reactions.
• (3)Water injection lowers combustion temperatures, reducing NOx emissions, but it increases unburned THC emissions by slowing the oxidation reaction rates. Minor effects were seen on CO emissions. Particle number decreased with water injection, especially the smaller nucleation mode particles.
• (4)Increasing the compression ratio further improves water injection benefits on specific fuel consumption. However, a higher compression ratio leads to more NOx emissions at higher loads. Combined application with a higher water-to-fuel ratio above 0.2 and higher compression ratio may be beneficial for fuel economics, especially for lower and middle loads.

The authors declare no competing financial interest.