Experimental analysis of the effect of hydrogen as the main fuel on the performance and emissions of a modified compression ignition engine with water injection and compression ratio reduction

Applied Thermal Engineering 238 (2024) 121933 Available online 7 November 2023 1359-4311/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Research Paper Experimental analysis of the effect of hydrogen as the main fuel on the performance and emissions of a modified compression ignition engine with water injection and compression ratio reduction J.M. Rueda-V´ azquez a , J. Serrano b , * , F.J. Jim´ enez-Espadafor b , M.P. Dorado a a Department of Physical Chemistry and Applied Thermodynamics, C´ ordoba University, Spain, Campus de Rabanales, Campus de Excelencia Internacional Agroalimentario ceiA3, 14071 C´ ordoba, Spain b Department of Energy Engineering, Seville University, Spain, Camino de los Descubrimientos, s/n, 41092 Sevilla, Spain ARTICLE INFO Keywords: Renewable fuel CO 2 reduction NO x emissions reduction High efficiency ICE ABSTRACT This study aimed to investigate the combined effects of compression ratio (CR), start of injection angle (SOI) and water injection (WI) on the performance, including NOx emissions, of a modified turbocharged compression ignition engine using H 2 as the primary fuel and diesel solely as a pilot ignition source. WI and H 2 were both injected into the intake manifold just after the charge heat exchanger. Test conditions varied the engine speed from 1500 rpm to 2750 rpm, covering a wide range of the original engine’s operating conditions, water mass flow varied from 0 to 0.795 g/cycle and SOI for pilot ignition was studied in the range 10◦, 5◦, 0◦before top dead center. Subsequently, the study examined the combined effect of reducing the engine’s CR from 17.5:1 (base) to 13.5:1, in one-point increments, reducing SOI and water addition to further enhance the hydrogen energy share (HES). HES varied from 0 % to 85 % under different engine conditions, and WI and SOI were employed to prevent knocking and ensure that the maximum combustion pressure did not exceed 160 bar. The test results indicate that NOx emissions increase with HES, but reducing the CR results in a reduction in NOx emissions of approximately 50 %. Additionally, WI further reduces NOx emissions by up to 50 %. The experimental evidence demonstrates that combining WI with CR and SOI reduction can effectively lower NOx emissions, achieving an efficiency greater than 35 % and therefore approaching the base diesel mode’s efficiency. This also reduces the maximum in-cylinder pressure, allowing for an increase in turbocharger pressure and consequently enhancing the engine’s specific power. The maximum engine power achieved was 31.6 kW at 2500 rpm (mean effective pressure of 10.1 bar), with a CR of 15.5, SOI at 10◦, 1.5 kg/kWh of water, a HES of 81 %, and a minimum NOx level of approximately 480 ppm, with a maximum combustion pressure below 120 bar. 1. Introduction To encourage the adoption of alternatives to fossil fuels and reduce greenhouse gas (GHG) emissions from hydrocarbon-based fuels in internal combustion engines in maritime and road transport, as well as electricity generation, governments have implemented stringent emission regulations [1,2]. Regarding road transport, the European Commission (EC) has proposed a 30 % reduction in CO 2 emissions limits for vehicles compared to the 2021 target, which was set at 95 g of CO 2 per kilometer [3]. Additionally, EC law now dictates that only power plants emitting less than 100 g of CO 2 /kWh can be considered ’environmentally sustainable’ [4,5]. These emission thresholds align with the Paris Agreement and are in line with the EU’s scientific objective to achieve carbon neutrality by 2050. In this regard, extensive efforts are being made to search for alternative fuels for diesel engines, such as oxygenated fuels like alcohol fuels, dimethyl ether, biodiesel fuels, and hydrogen as the most promising options. Biodiesel is a viable solution for diesel engines due to its biodegradable, oxygenated, sulfur-free, and renewable characteristics [6]. Blending biodiesel with fossil diesel offers several benefits, including reduced emissions, decreased engine wear, and comparable thermal efficiency compared to pure diesel fuel [7]. It’s worth mentioning blends like diesel-tung oil-ethanol [8] and diesel-palm oilethanol [9]. In both cases, the feasibility of these blends as diesel fuel * Corresponding author. E-mail address: [email protected] (J. Serrano). Contents lists available at ScienceDirect Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng https://doi.org/10.1016/j.applthermaleng.2023.121933 Received 5 August 2023; Received in revised form 9 October 2023; Accepted 6 November 2023 Applied Thermal Engineering 238 (2024) 121933 2 alternatives has been demonstrated using current diesel technology. Furthermore, extensive experimental work has been carried out to analyze the suitability of waste cooking oil (WCO) [10], where Joonsik Hwang and colleagues compared the performance of this fuel, highlighting a reduction in soot and CO emissions and a slight increase in indicated specific fuel consumption. However, biodiesel has significant limitations as a complete alternative to petroleum derivatives due to several factors, such as elevated production costs, feedstock limitations related to competition with food production, the potential for deforestation, environmental concerns, and relevant regulatory and policy issues. Therefore, the quest for sustainable and environmentally friendly energy solutions has sparked renewed interest in hydrogen as a clean alternative to conventional fossil fuels. Among the various applications of hydrogen as an energy carrier, internal combustion engines (ICEs) show significant promise due to their adaptability to existing infrastructure and their potential to facilitate the transition to a hydrogenbased economy. When coupled with innovative technologies that offer flexibility to meet the engine’s required characteristics [11,12], H 2 offers several advantages: 1. Zero carbon content (despite not existing freely in nature, hydrogen is the most abundant element in the universe) [13]. 2. It can be produced from renewable energy sources, although currently, 95 % of hydrogen is generated using methods based on fossil fuels [14]. 3. It possesses a very high energy density of 120 MJ/kg, which is around 3 times that of hydrocarbon-based fuel. 4. Extensive research is currently underway for both hydrogen production and storage [12,13]. However, it is worth noting that when it comes to hydrogen storage, there are some drawbacks, such as its low volumetric energy density, which necessitates large storage areas and makes it expensive. Compressed H 2 storage is the most established hydrogen storage technology and involves pressurizing H 2 in high-pressure vessels; with the current technology based on wrapped composite cylinders, pressures of up to 1000 bar can be achieved [15]. However, there are new exploration areas, such as electrochemical hydrogen storage materials in nanostructures [16,17]. In these cases, researchers are seeking materials with a large surface area as well as proper porosity for H 2 storage [18,19], which appear to have significant potential. Using H 2 as a fuel in an internal combustion engine (ICE) involves significantly different combustion properties compared to hydrocarbonbased fuels. Hydrogen burns much faster than these fuels, which means that achieving high efficiency with H 2 in an ICE requires optimizing the shape of the combustion chamber, adjusting the CR, and calibrating the ignition timing to prevent damaging knock [20], minimize elevated NOx emissions [21,22], reduce high in-cylinder peak pressures and temperatures, and improve engine performance [23]. On the other hand, the high diffusivity of hydrogen allows for a perfect mixture with air, achieving almost complete homogeneity when introduced into the combustion chamber. Although H 2 can be used in both spark ignition (SI) [24,25] and compression ignition engines (CI) [26], it cannot be directly used in CI engines due to its higher selfignition temperature (850 K). This would require an impractically high CR or heating the intake air, which ultimately reduces volumetric efficiency and engine power. However, H 2 can be utilized in CI engines in dual-fuel mode by using diesel fuel as an ignition source. Typically, gaseous H 2 fuel is supplied to the intake manifold, and diesel fuel is directly injected into the combustion chamber as the ignition source. However, this approach results in reduced volumetric efficiency since H 2 displaces part of the intake air during the intake stroke [27]. There are two drawbacks related to using H 2 as fuel in CI engines: NOx generation and knocking, which prevent achieving maximum power. In this regard, there are several strategies for controlling them, such as exhaust gas recirculation (EGR), direct water injection (DWI), or water injection into the intake manifold (WI) [28]. All of these strategies aim to reduce in-cylinder temperatures, primarily through the Zeldovich mechanism, which is one of the main mechanisms that contribute to NOx formation. The addition of water has various effects on combustion, including a decrease in in-cylinder temperature, resulting from the increased heat capacity of the mixture (thermal effect); a delay in the SOC due to reduced load temperature; a minor chemical effect on engine combustion due to OH radicals from water [29]; and a decrease in oxygen concentration (dilution effect), which leads to reduced NOx emissions. Nomenclature ATDC After top dead centre BTDC Before top dead centre CFD Computational Fluid Dynamics CI Compression ignition engine CO Carbon monoxide CO 2 Carbon dioxide C p Specific heat at constant pressure CR Compression ratio DWI Direct water injection EC European Commission ECU Electronic control unit EGR Exhaust gas recirculation FB Flow-blurring FSN Filter smoke number GHG Greenhouse gases H 2 Hydrogen H 2 O Water HES Hydrogen energy share HRR Heat release rate ICE Internal combustion engine MEP Mean effective pressure N 2 Nitrogen NO Nitric oxide NO 2 Nitrogen dioxide NOx Nitrogen oxides OH Hydroxyl radical PM Particulate matter RPM Revolutions per minute SI Spark ignition SM Single mode SML Single mode with partial low rate SOC Start of combustion SOI Start of injection T Temperature TDC Top dead centre TM Two modes WCO Waste cooking oil WI Water injection into the intake manifold Greek symbols α Ratio between thermal capacity of the engine charge and fuel, both in terms of power η Efficiency Δ Delta increment J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 3 Another alternative for controlling NOx emissions and knocking and addressing the challenges associated with using H 2 as a fuel, is the reduction of the CR in CI engines. Masood et al. [30] analysed CR reduction in dual-fuel engines (H 2 -diesel) and reported an increase in the amount of H 2 that can be used. However, the efficiency of the engine in dual-fuel mode decreases with CR reduction due to reduced combustion efficiency and lower in-cylinder temperatures resulting from the lower CR [31,32]. Another focus is provided by V. Chintala et al. [22], where CR reduction and WI operated in a CI under dual-fuel mode allowed a significant reduction in NOx emissions compared to the base diesel mode but slightly higher than the conventional dual-fuel mode. A notable conclusion emerging from this study is that the combined strategy of water addition and CR reduction is a promising solution for a drastic enhancement of the maximum hydrogen energy share. This paper presents the results of an experimental investigation into the combustion characteristics and NOx emissions of a hydrogen-fueled internal combustion engine equipped with a water injection system. The study focuses on exploring the influence of hydrogen concentration, water injection volume, and CR on MEP, NOx emissions, ignition delay, maximum in-cylinder pressure, and efficiency under various engine loads and speeds. The combination of hydrogen as the primary fuel and water injection as a control strategy has the potential to enhance combustion stability, mitigate knock risk, decrease peak temperatures, and reduce NOx formation during the combustion process. The analysis carried out in this work regarding the combined effects of CR, SOI, WI, and HES on combustion and performance at four different RPM settings in a four-cylinder turbocharged diesel engine has not been undertaken by other researchers. This is evident from the review by Pali Rosha et al. [33] and the research conducted by V. Chintala [34], enabling the projection of the study’s results to a practical implementation. To gain a comprehensive understanding of the current state of technology regarding the use of H 2 as fuel in CI engines and facilitate a comparison with the analysis conducted in this work, Table 1 presents a summary of the primary experimental studies currently published. As can be seen in Table 1 and as will be discussed later, this work covers a broader range of parameters: - The minimum CR studied in this work is 13.5:1, while other researchers have either focused on a single CR or studied down to 15.4:1. [22,30,34,35,38,39]. Table 1 Summary of the primary experimental studies currently published on H 2 dual fuel CI engines with H 2. Author (reference) H 2 enrichment and mixing process Engine specifications WI/CR tested Key findings Works with both WI and CR reduction Chintala and Subramanian [34] 32 % HES (Manifold injection) Single cylinder Max Power: 7.4 kW at 1500 rpm WI 130–410 g/ kWh CR 19.5:1 16.5:115.4:1 NOx emissions decrease with higher WI rates. (CR 19.5:1, 130 g/kWh WI) NOx =9.7 g/kWh (CR 19.5:1, 270 g/kWh WI) NOx =6.5 g/kWh (CR 19.5:1, 410 g/kWh WI) NOx =5.1 g/kWh Max. in-cylinder temperature decrease with higher WI rates. (CR 19.5:1, 340 g/kWh WI) T =1784 K (CR 16.5:1, 340 g/kWh WI) T =1437 K (CR 15.4:1, 340 g/kWh WI) T =1248 K Raju and Masimalai [35] 8 % HES (Manifold injection) (Numerical study) Single cylinder Max Power: 3.67 kW at 1500 rpm WI 3–7 % by mass CR 16.5:115.5:1 Results predict a rise in in-cylinder temperature at 8 % HES, which also increases knocking and NOx emissions. The higher in-cylinder temperature as well as its effects are reduced by the use of WI and also by reducing the CR in parallel. Works with WI Chintala and Subramanian [36] 0–39 % HES (Manifoldinjection) Single cylinder Max Power: 7.4 kW at 1500 rpm WI 0–270 g/kWh NOx emissions increase with higher HES. NOx emissions are reduced a maximum of 37 % with 270 g/kWh WI, with the penalty of a slight increase in carbon-based emissions. Chintala and Subramanian [37] 18 % HES (Manifoldinjection) Single cylinder Max Power: 7.4 kW at 1500 rpm WI 200 g/kWh NOx emissions increase with higher HES. (No H 2 , No WI) NOx =6.9 g/kWh (19 % HES, No WI) NOx =9.7 g/kWh (19 % HES, 200 g/kWh WI) NOx =6.6 g/kWh Serrano et al [29] 0–74 % HES (Manifold injection) Four cylinders Max. Power: 63 kW at 4000 rpm WI 0–48 kg/h NOx emissions increase with higher HES. NOx emissions decrease with higher WI rates. WI from 0 to 16 kg/h produces a reduction in NOx emissions of approximately 70 %, compared to the cases without WI.The thermal mechanism clearly dominates and provides a minimum value for NOx emission with regard to the thermal capacity of the engine load. Works with CR reduction Chintala and Subramanian [22] 0–63 % HES (Manifold injection) Single cylinder Max Power: 7.4 kW at 1500 rpm CR 19.5:1 16.5:115.4:1 NOx emissions decrease with lower CRs. NOx emissions increase with higher HES. (CR 19.5:1, 19 % HES) NOx =9.7 g/kWh (CR 16.5:1, 19 % HES) NOx =5.5 g/kWh (CR 15.4:1, 19 % HES) NOx =5 g/kWh Masood et al [30] 8.3–30.1 % HES (Direct Injection) Single cylinder Max. Power: 3.7 kW at 1500 rpm CR 24.5:1 22:1 20:1 18.35:116.35:1 NOx emissions decrease with lower CRs. NOx emissions increase with higher HES. (CR 24.5:1, 90 % HES) NOx =38 % higher than 10 % HES. (CR 18.35:1, 90 % HES) NOx =27 % higher than 10 % HES. (CR 24.5:1, 90 % HES) NOx =38 % higher than CR 16.35:1. Rosha et al [38] 0.5 kg/h (H 2 -enriched biogas) (Manifold Injection) Single cylinder Max. Power: 3.5 kW at 1500 rpm CR 18:1 17:116:1 NOx emissions decrease with lower CRs. (CR 18:1, 3.5 bar MEP) NOx =16.5 g/kWh (CR 17:1, 3.5 bar MEP) NOx =13.2 g/kWh (CR 16:1, 3.5 bar MEP) NOx =12.6 g/kWh Yilmaz [39] 0–37.6 % HES (Manifold Injection) Four cylinders Max Power: 48 kW (4000 rpm) CR 18.25:116.9:1 H 2 enrichment and low CR can be used to reduce the maximum in-cylinder pressure and HRR and prevent abnormal combustion without deteriorating the BTE. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 4 Fig. 1. (a) Schematic diagram of the experimental setup. (b) Photograph of the experimental test rig. (c) Schematic diagram of the FB atomizers. (d) Schematic diagram of the H 2 supply and water addition system. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 5 - In this work, the range of HES has been continuous, spanning from 0 to 85 %, while previous studies have focused on specific values of HES or reduced ranges [22,30,29,34,35,36,37]. - The combination of WI and CR in a current technology engine, turbocharged and with 4 cylinders, has only been studied in this work; the rest of experimental analyses involving CR and WI have been conducted on naturally aspirated single-cylinder engines [29]. - The analysis performed in this work has covered a range from 1500 rpm to 2750 rpm, whereas results provided by other researchers have been reported at only one RPM [22,34,35,36,37,38,39]. - The maximum MEP achieved with H 2 (with diesel fuel only for ignition) is the same as that of the original diesel engine at maximum power. There is no published work where the performance has been analyzed at this MEP. The objectives of this research are as follows: 1. To evaluate the influence of water injection on combustion characteristics and efficiency in a hydrogen-powered ICE at varying CR, while relying solely on diesel fuel for ignition, over a wide range of CR values. 2. To quantify the reduction in NOx emissions achieved by employing WI, varying SOI, and adjusting CR across a broad range of RPMs, all while maximizing HES. 3. To assess the trade-offs between performance and emissions control in a hydrogen-powered ICE by examining the effects of varying compression ratios, SOI timing, and WI strategies. 2. Experimental setup The experimental analysis conducted in this study was performed on a commercial diesel engine that was modified to enable the measurement of combustion pressure, utilize hydrogen as a fuel source, and introduce water injection into the intake manifold (refer to Fig. 1). The original engine is the Renault model 1.5 dCi k9k 766, and its key characteristics are detailed in Table 2. In this research, we investigated five different compression ratios: 17.5:1, 16.3:1, 15.5:1, 14.5:1, and 13.5:1. These ratios were achieved by modifying the piston head shape. Commercial solutions exist for 17.5:1 and 15.5:1, as illustrated in Fig. 2 (a) and (c). However, for 16.3:1, 14.5:1, and 13.5:1, grooves of varying depths were created to remove material of from the piston head, demonstrated in Fig. 2(b), (d), and (e). Specifically, for 16.3:1, the grooves were incorporated into the 17.5:1 commercial piston, while for 14.5:1 and 13.5:1, they were implemented in the 15.5:1 commercial piston. Table 3 lists all the sensors installed in the engine system depicted in Fig. 1. The temperature and pressure sensors are denoted by the letters “T” and “P,” respectively, while the ammeter is represented by the letter “A.” The P2 and P3 pressure sensors are “ThermoCOMP™ Quartz” water-cooled measuring probes designed with compact dimensions, particularly suitable for use in internal combustion engines with reduced volume and more than 2 valves per cylinder. The pressure sensor model is 6043A60 by Kistler [40]. The ammeter measures the current signal of the injector, enabling simultaneous measurement of the injection duration and the start of injection angle. The engine configuration and timing were adjusted using the Commander 4DE from Sybelle, which is the electronic control unit (ECU) capable of modifying most of the engine parameters [41]. In the context of this study, this unit was utilized to manage the following engine subsystems: the start of injection angle (SOI), the quantity of fuel for preinjections, and the achievable turbocharger pressure (via a map correlating RPM and injection quantity). Although the engine was equipped with an exhaust gas recirculation valve (EGR), it was deactivated for this research. In this study, NOx exhaust gas emissions were measured using a Testo 350 XL. The opacity of the exhaust gas was assessed with an AVL Smoke Meter 415SE, which employed the filter paper method to determine the filter smoke number (FSN), as defined by ISO 10054. 3. Experimental tests The engine running on hydrogen was studied examined at four different speeds (1500, 2000, 2500, and 2750 rpm) and various SOI settings (+10◦, +5◦, and 0◦BTDC). For each SOI setting, the amount of diesel fuel injected was adjusted to achieve the target speed (without hydrogen) and no without any applied brake load. The engine’s wastegate control was deactivated, resulting in the intake pressure being set to the maximum attainable level based on the exhaust conditions at the turbine inlet. The tests were conducted for each CR as follows: at each engine RPM and SOI combination, the mass flow of hydrogen was incrementally increased gradually raising engine torque and as observed, leading to higher NOx levels. When the NOx emissions reached a certain threshold or if knocking was detected, water was injected using FB atomizers (to minimize water coalescence), effectively reducing NOx levels but affecting the HRR and overall engine performance. The test continued until one of the following limits was reached, although both typically met occurred simultaneously: - The maximum allowable in-cylinder pressure to ensure mechanical integrity is approximately 150–160 bar. - Combustion knocking. The annex shows the various tested conditions, with over 469 tests conducted, and provides the minimum and maximum values for the following variables: hydrogen mass flow, water mass flow, diesel mass flow, torque, power, MEP, efficiency, and the (diesel +H 2 )/air ratio. A positive main injection angle indicates that the SOI occurs before TDC. The test conditions allow for the investigation of the interaction between H 2 and water, as well as the efficiency and the fundamentals of NOx formation. Furthermore, these test conditions have enabled the attainment of a mean effective pressure of 11.3 bar, which represents 90 % of the mean effective pressure of the diesel engine at maximum power (12.6 bar). This allows for a meaningful comparison between the hydrogen-based solution and the diesel engine. In this experimental study, in-cylinder temperature and HRR were determined using a 0-D thermodynamic model developed by J. Serrano et al., as cited in Ref. [29]. As outlined in Section 2, modifications were made to the piston heads to achieve the desired CR, resulting in changes in bowl shape and the cylinder/bowl ratio for each CR. While these modifications may influence mixture formation and diesel combustion, the minimal quantity of diesel injected and its correspondingly short duration ensure that the combustion chamber’s shape does not significantly impact diesel combustion or the SOC, as illustrated below. Figs. 3 and 4 depict the injection pulse intensity in amperes, as Table 2 1.5dCi k9k 766 Engine characteristics in diesel-only mode. Type 4 S, 4 cylinders Engine displacement 1461 cm 3 Bore 76 mm Stroke 80 mm Compression ratio 17.5:1 Maximum power 63 kW at 4000 rpm Maximum torque 220 Nm at 1800 rpm Fuel Diesel EN590 Lubrication Gear oil pump Injection pump Distributor injection pump with common rail Injector Solenoid type EGR Valve controlled and water cooled Turbocharger Turbine controlled by waste gate valve Engine refrigeration Water cooled Charge air Water cooled J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 6 measured by the ammeter in Fig. 1a, along with the combustion pressure and the HRR curve, which enables the assessment of the SOC. It is evident that, in all conditions, the SOC occurs after the conclusion of injection. As a result, all of the diesel fuel is inside the combustion chamber before the combustion process initiates, ensuring an ample supply of air for the fuel–air mixture. Under these circumstances, the SOC is solely dependent on the pressure–temperature history of the fuel [42] since the SOI, with no significant influence from the combustion chamber shape. The opposite scenario would be applicable to typical diffusive combustion when the entire range of engine power is solely provided by diesel. The smallest difference between the end of injection and the SOC occurs at the highest CR, mainly due to the higher charge temperature associated with this CR. It is also noticeable at the highest engine speed, primarily because of the increased quantity of diesel injection, particularly when water injection is not utilized. This is illustrated in Fig. 4(a), where, under these conditions, the injection concludes just before the SOC, as confirmed by the comparison between the injection pulse and Fig. 2. Photographs of the different combustion chamber. (a) CR =17.5, (b) CR =16.3, (c) CR =15.5, (d) CR =14.5 and (e) CR =13.5. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 7 the HRR. WI also results in a delay in the SOC, as demonstrated in Fig. 3(b) and Fig. 4(b) (this aspect will be further discussed in Section 4.3). In these figures, a positive crank angle value indicates “after Top Dead Center” (ATDC), while a negative value indicates “before Top Dead Center” (BTDC). When comparing the injection duration between the lowest engine running speed (Fig. 3) and the highest engine speed (Fig. 4), it becomes evident that the injection pulse lasts longer (in terms of crank angle) in the latter case due to the increased pilot fuel quantity at higher engine speeds, as specified in the annex. This extended duration is necessary to counteract the rising friction losses associated with higher engine speeds. 4. Results and discussion - performance and emissions with different compression ratios The analysis of the experiments focuses on four specific aspects: NOx emissions, engine efficiency, SOC analysis, and maximum combustion pressure. In hydrogen-based fuel engines, increasing the hydrogen energy share (HES) also results in an increase in NOx emissions due to faster combustion resulting from the high diffusivity and flame speed of hydrogen. To mitigate this undesirable effect, various strategies can be employed: modifying the SOI, enabling exhaust gas recirculation (EGR, which is not studied in this work), implementing WI, and reducing the CR. These strategies could offer solutions to enhance the maximum HES, leading to increased engine specific power, reduced NOx emissions, improved engine efficiency, and, for these reasons, the effects of CR, SOI angle, and WI will be studied in this section through the analysis of MEP, efficiency, SOC, HRR, and in-cylinder maximum pressure. The HES is evaluated using the dimensionless Equation (1), and in the experiments conducted in the annex, it falls within the range of 48 % to 85 %: HES =mH2•LHVH2 mH2•LHVH2+mdiesel •LHVdiesel (1) All the figures presented in this paper follow the same legend structure. Each colour and symbol represent varying values for H 2 O mass flow, measured in grams per cycle, and different levels of HES. Regarding SOC, this can be seen as a contributing factor in explaining the effects of H 2 and WI. Although SOI is the independent variable, SOC is the one that is truly important for understanding the impact of SOI on emissions and efficiency. Furthermore, SOC is sensitive to CR, HES, and WI because it exhibits a high dependency on charge temperature. Hydrogen combustion generates only water and does not produce any PM (soot). Soot originates from the diesel fuel used for ignition in the combustion process and consistently remains at low levels in all tests, below 2 mg/m 3 . For this reason, it has not been included in the analysis. 4.1. No x emissions analysis The formation of NOx is primarily explained by the Zeldovich mechanism, which exhibits a strong dependence on temperature and a lesser dependence on the fuel–air equivalence ratio. NOx is generated through the oxidation of N 2 present in the inducted air, driven by the high in-cylinder temperature, and is primarily composed of NO with smaller amounts of NO 2 . The rapid combustion of H 2 tends to elevate the gas temperature on in the combustion chamber, thereby promoting the formation of NOx. In this study, under identical operational conditions (RPM, SOI, HES, and H 2 O mass flow), it is observed, at as expected, that NOx emissions decrease as the CR decreases. Figs. 5 and 6 illustrate the evolution of NOx emissions versus CR at different SOI and RPM, while considering HES and WI as parameters. Meanwhile, Fig. 7 depicts the calculated mean combustion temperature of the gases within the combustion chamber versus crank angle, considering different HES, RPM, and SOI, and utilizing CR and WI as parameters. The dependence of NOx on RPM is clearly evident for all the tested conditions and the five different CR values. When examining Fig. 5(a), (b), and (c), all of which have the same SOI of 10◦, the maximum NOx concentration decreases progressively from approximately 1000 ppm at 2000 rpm to below 600 ppm at 2750 rpm. In the case of a 5◦SOI, as seen in Fig. 6(a) and (b), NOx decreases from a peak of 800 ppm at 1500 rpm to 300 ppm at 2750 rpm, with nearly the same HES. This trend across RPM, which applies to all conditions, is attributed to the kinetics of NOx formation. Considering that the combustion process occurs within a crank angle range of 5◦to 25◦, as engine RPM increases, the available time for complete NOx formation diminishes, resulting in less time for NOx to fully develop. Therefore, from this perspective, the higher the RPM, the lower the expected NOx formation. The dependence of NOx on SOI can be observed across the tested conditions and the five different CR values. When comparing Fig. 5(a), 6 (a), and (c), where the SOI is 10◦, 5◦, and 0◦respectively, the maximum NOx concentration progressively decreases from approximately 1000 ppm at SOI 10◦to roughly 300 ppm at SOI 0◦. Additionally, examining Fig. 5(c) and 6(b), both at 2750 rpm and with a SOI of 10◦and 5◦, respectively, NOx decreases from 600 ppm to 300 ppm. This behaviour is attributed to the significant dependence of NOx formation on temperature. Once combustion begins, NOx forms at the flame front (which is very narrow due to the high pressure inside the combustion chamber), but as pressure and temperature continue to increase, NOx continues to form in the burned area. When SOI is reduced, there is also a reduction in SOC (achieve negative values), as can be seen when comparing Fig. 7 (a) and (b). Consequently, the portion of combustion occurring at higher pressure and temperature decreases, leading to a reduction in NOx generation. This effect is clearly evident when examining Fig. 7(a) and (b), both at 2000 rpm, with SOIs of 10◦and 0◦, respectively. The consistent temperature reduction with decreasing SOI aligns with the previously mentioned trend. In particular, it can be observed that at CR 17.5 and with approximately the same HES, the maximum temperature decreases by 200 ◦C. With a reduced CR, the loss in maximum temperature is around 100 ◦C. The dependence of NOx on HES can be observed when we exclude the effects of WI; therefore, only the cases with 0 g/cycle of water are considered here. When examining Fig. 5(a) (2000 rpm and SOI 10◦) and comparing HES levels of 45 % and 62 %, NOx increases from 420 to 900 Table 3 Sensors used in the experimental apparatus. Temperature sensors Position Comments Unit Measurement accuracy T1 Mass air flow ◦C (-40 +120) ◦C (≤3 %) T2 Intake air before compressor 0–1000 ◦C ±1.5 ◦C T3 Intake air after compressor T4 Intake air after cooler T5 Intake air after water injection T6 Exhaust flow before turbine T7 Exhaust flow after turbine Pressure sensors Position Comments Unit Measurement accuracy P1 Outlet compressor pressure bar (absolute) 0–10 bar (≤0.5 %) P2 Combustion pressure inside cylinder bar (relative) 0–250 bar ±0.5 % P3 Current sensor Position Comments Unit Measurement accuracy A1 Ammeter A 0–30 A ±1 % J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 8 ppm, respectively. In Fig. 5(b) (2500 rpm and SOI 10◦), NOx emissions continuously rise from 340 ppm at 38 % HES to 810 ppm at 64 % HES in an almost quadratic fashion. In Fig. 5(c) at 2750 rpm (the highest tested) and SOI 10◦, NOx emissions increase from 240 ppm at 34 % HES to 590 ppm at 61 % HES. This continuous increase in NOx with HES is attributed to the rise in combustion gas temperature resulting from the increased presence of H 2 as engine load increases. The temperature dependence can be observed in Fig. 7(d), where the mean combustion temperature clearly increases with HES for the same RPM. The effect of WI on NOx is clearly evident in almost all the Figs. 5 and 6. In fact, a noticeable reduction in NOx is observed with each increase in WI. In Fig. 5(a), with a CR of 17.5 and an HES of 62 %, NOx decreases from 900 ppm without WI to 380 ppm with 0.266 g/cycle of water. This means that water has halved the emission levels. Additionally, in the same figure, at CR 17.5, NOx decreases from 620 ppm at HES 72 % with 0.266 g/cycle of water to 450 ppm at HES 77 % with 0.531 g/cycle of water. In this case, NOx reduces by 50 % with an almost 100 % increase in WI, despite a 7 % increase in HES. Lastly, in the same figure, it can be seen that at CR 16.3 and HES 80 %, NOx emissions decrease from 1050 ppm with 0.531 g/cycle of water to 700 ppm with 0.795 g/cycle of water. In this case, a 33 % reduction in NOx has been achieved with a 50 % increase in WI. A very similar behaviour can be appreciated in the rest of tested conditions. The procedure by which WI reduces NOx is named the thermal mechanism, where two different mechanisms occur: the first one is the increase in the specific heat of the gases, which reduces the charge temperature during compression and combustion due to the high thermal capacity of water compared to that of the air. The second one is related to the phase change of water; most of the water that enters the cylinder is in a liquid state, although highly atomized. As the compression process advances and the temperature increases, the water undergoes a phase change from a liquid to a gas state, and therefore, it absorbs the energy required for the phase change. The effect of WI on NOx can also be observed in Fig. 7, which Fig. 3. (a) Comparison of the injection pulse duration and the SOC through the HRR at 1500 rpm and SOI 10 ◦for CR 17.5, HES 54 % and 0.0 g/cycle H 2 O. (b) Comparison of the injection pulse duration and the SOC through the HRR at 1500 rpm and SOI 10◦for CR 17.5, HES 54 % and 0.355 g/cycle H 2 O. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 9 displays the in-cylinder temperature curves for most of the test conditions presented in Fig. 5 and Fig. 6, as modelled by J. Serrano et al. in references [38,39]. Looking at Fig. 7(a) and (b), it is evident that for the same HES, SOI, and CR, any increase in the amount of WI leads to a rapid reduction in temperature evolution and, consequently, in NOx emissions. However, this effect is more pronounced at SOI 5◦than at SOI 10◦. To put it differently, the higher the SOI, the greater the amount of water required for the same temperature reduction. To emphasize this point, in Fig. 7(a) at HES 62 %, SOI 10◦, and CR 16.3, an increase from 0 to 0.266 g/cycle of water results in a reduction of the maximum temperature by almost 70 ◦C. In Fig. 7(b), at HES 64 %, SOI 5◦, and CR 16.3, increasing from 0 to 0.213 g/cycle of water reduces the maximum temperature by almost 120 ◦C. This is because, with a lower SOI, the maximum temperature inside the combustion chamber is lower, and therefore, with the same thermal capacity of the water (amount of WI per engine cycle), the reduction in temperature is more significant. Regarding the dependence of NOx on CR, it can be observed in Figs. 5 and 6 that there is an almost linear relationship between CR and NOx. Therefore, the reduction of NOx can be promoted by decreasing CR, regardless of RPM, SOI, or HES. This is because any increase in CR has two effects: it raises the air temperature and shortens the start of combustion (precisely due to the temperature increase). Therefore, reducing CR lowers the gas temperature, deactivating NOx formation. Additionally, the amount of water required to achieve a certain level of NOx is also reduced. If the effects of both RPM and SOI on NOx formation are analysed for the same amount of H 2 supplied to the engine, a significant reduction in NOx can be observed. When comparing Fig. 5(a) and 6(b) for approximately 20 kW of H 2 and a CR of 16.3, with engine speed ranging from 2000 rpm to 2750 rpm and SOI decreasing from 10◦to 5◦, the NOx levels decrease from 1050 ppm to 140 ppm, with the latter having a 27 % lower WI. This trend can be observed in the remaining tested conditions, highlighting the cumulative effect of RPM and SOI on NOx reduction. An analysis of the test data determined the limit for the minimum Fig. 4. (a) Comparison of the injection pulse duration and the SOC through the HRR at 2750 rpm and SOI 10 ◦for CR 17.5, HES 61 % and 0.0 g/cycle H 2 O. (b) Comparison of the injection pulse duration and the SOC through the HRR at 2750 rpm and SOI 10◦for CR 17.5, HES 61 % and 0.194 g/cycle H 2 O. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 16 ultimately, SOC is delayed with higher RPM. This trend is clearly observable at SOI 5◦in Fig. 11(b), (c), and (d), where under all CR and test conditions, SOC consistently increases with RPM. To provide clarity, let’s focus on the test condition without WI and with the same HES (represented by the dark blue line). At a CR of 13.5, SOC increases from 5◦at 1500 rpm to 8.3◦at 2750 rpm. When the CR is increased to 17.5, SOC increases from −0.3◦at 1500 rpm to 3◦at 2750 rpm. This behaviour is also evident when comparing Fig. 11(a) and (e) for SOI 10◦. The dependence of SOC on HES can be observed when we exclude the effects of WI; therefore, only the cases with 0 g/cycle of water are considered here. When examining Fig. 11(c) (2500 rpm and SOI 5◦) and comparing HES levels of 38 %, 55 %, and 64 % for CR 13.5, SOC decreases from 7◦to 6◦to 5◦, respectively. In the same Fig. 11(c), for CR 15.5, SOC steadily decreases from 5◦at 38 % HES to 2.2◦at 64 % HES in an almost linear fashion. Finally, in the same Fig. 11(c) at CR 17.5, SOC goes from 3.5◦to 2.1◦to 1.0◦for HES 38 %, 55 %, and 64 %, respectively. Fig. 14 displays the HRR curves for most of the test conditions presented in Fig. 11, as modeled by J. Serrano et al. in references [38,39]. In this Fig. 14(a) (CR 17.5 and 2000 rpm), it is clearly demonstrated that an increase in HES results in a reduction in SOC. Firstly, the area of HRR increases with HES because of the increased amount of H 2 , and at the same time, the maximum of the HRR also rises, raising gas temperature. This is the main factor responsible for SOC reduction; as temperature increases within the combustion pressure range, the kinetics of the fuel that produces autoignition accelerates. Therefore, this continuous reduction in SOC with HES is attributed to the increase in combustion Fig. 10. (a) MEP versus CR at 2000 rpm and SOI 10 ◦BTDC. (b) MEP versus CR at 2500 rpm and SOI 10◦BTDC. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 17 Fig. 11. (a) SOC versus CR at 2500 rpm and SOI 10 ◦BTDC, (b) SOC versus CR at 2750 rpm and SOI 5◦BTDC, (c) SOC versus CR at 2500 rpm and SOI 5◦BTDC, (d) SOC versus CR at 1500 rpm and SOI 5◦BTDC, (e) SOC versus CR at 2750 rpm and SOI 10◦BTDC. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 18 gas temperature due to the greater presence of H 2 as engine load increases at the same RPM. Although the specific heat at constant pressure of H 2 is significantly higher compared to that of air roughly fourteen times greater, to be precise the low mass of H 2 leads to a continuous increase in gas temperature with an increase in engine load. The effect of WI on SOC can be observed in almost all of Fig. 11(a)– (e). For the same HES, the higher the WI, the higher the SOC. This is due to the temperature reduction caused by the presence of water in the gases, which delays the start of combustion. In Fig. 11(a) (2500 rpm and SOI 10◦) with HES 64 % and CR 13.5, SOC increases from 0◦without WI to 3◦with WI of 0.213 g/cycle of water. At CR 17.5 and HES 64 %, SOC changes from −8◦without WI to −3◦with 0.213 g/cycle of water. When looking at HRR, the dependence of SOC on WI is clearly evident in a physical context. In Fig. 14(b) (CR 16.3, 2000 rpm and SOI 10◦) at HES 77 %, SOC goes from −3.6◦to −3.0◦as WI increases from 0.266 g/cycle to 0.531 g/cycle, respectively. Moreover, at HES 80 %, the SOC changes from −4.3◦to −3.2◦with an increase in WI from 0.531 g/cycle to 0.795 g/cycle, respectively. In this Fig. 14(b), it can be seen that not only is SOC delayed as WI increases for the same HES, but the maximum of the HRR also shifts roughly by the same angle. In other words, the entire combustion process is delayed along with SOC. A very similar behaviour is observed in Fig. 14(c) (CR 15.5, 2000 rpm and SOI 10◦) at HES 77 %, where SOC goes from −1◦to 0◦as WI increases from 0.266 g/cycle to 0.531 g/cycle, respectively. Moreover, at HES 80 %, the SOC changes from −2◦to 3◦with an increase in WI from 0.531 g/cycle to 0.795 g/ cycle, respectively. As seen, the SOC is delayed by approximately 1◦for an increase in WI of around 100 %, although the initial angle depends on the CR. Regarding the dependence of SOC on CR, it can be observed from Fig. 11 that there is an almost inverse relationship between CR and SOC: the higher the CR, the lower the SOC. Therefore, an increase in CR results in a reduction of SOC, regardless of RPM, SOI, WI, or HES. This is because any increase in CR raises the air temperature and shortens the temporal history of the fuel that controls the start of combustion within the pressure range. 4.4. Effect on the in-cylinder maximum pressure The maximum pressure inside the combustion chamber is of great interest because it affects the mechanical integrity of the engine. As mentioned in section 3, the specific power of the engine is limited not only by knock but also by mechanical constraints. Therefore, in this Fig. 11. (continued). J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 19 section, the factors that control peak pressure in dual combustion are identified. Fig. 12(a) and (b) depict combustion pressure versus crankshaft angle for various testing conditions. Common observations from both figures are as follows: - The higher the HES, the greater the pressure at the end of compression and also the maximum pressure. This is attributed to the increased pressure ratio of the compressor with engine load, resulting in higher intake pressure. This trend is evident for any CR, WI, SOI, or RPM. - Increasing WI has only a marginal effect on reducing the maximum pressure. The pressure is more responsive to an increase in HES than to an increase in WI. To delve deeper into the subject, changes in the maximum incylinder pressure can be observed in Fig. 13(a) to (e). These figures illustrate the trends in maximum in-cylinder pressure with increasing HES, ranging from a CR of 13.5:1 to 17.5:1. For the same HES and H 2 O content, there is an observed increase of between 20 and 40 bar in pressure from CR =13.5 to CR =17.5. This trend holds true across all operational conditions and HES ranges. For instance, in Fig. 13(b), for the red line representing 72 % HES and 0.266 g/cycle H 2 O, the maximum pressure at CR =13.5 is 60 bar, while at CR =17.5, it reaches 100 bar. Conversely, in Fig. 13(d), for the orange line with 63 % HES and 0.0 g/cycle H 2 O, the maximum pressure at CR = 13.5 is 65 bar, and at CR =17.5, it reaches 85 bar. This indicates a continuous increase in maximum pressure with increasing CR. This behavior is valuable as it demonstrates that any CR increase exacerbates the mechanical load on the engine, which is contrary to the perspective of enhancing engine efficiency. WI also leads to a reduced drop in in-cylinder pressure. If you observe Fig. 13(b), there is a 5-bar difference between the yellow line (62 % HES and 0.0 g/cycle H 2 O) and the orange one (62 % HES and 0.266 g/cycle Fig. 12. Combustion pressure at different HES and WI. (a) CR 16.3, SOI 5 ◦and 2500 RPM, (b) CR 17.5, SOI 10◦and 2750 rpm. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 20 Fig. 13. (a) In-cylinder maximum pressure versus CR at 1500 rpm and SOI 10 ◦BTDC, (b) In-cylinder maximum pressure versus CR at 2000 rpm and SOI 10◦BTDC, (c) In-cylinder maximum pressure versus CR at 2000 rpm and SOI 5◦BTDC, (d) In-cylinder maximum pressure versus CR at 2500 rpm and SOI 5◦BTDC, (e) Incylinder maximum pressure versus CR at 2750 rpm and SOI 10◦BTDC. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 21 H 2 O). This trend is consistent across all CR values and operational conditions. A similar decrease in the pressure can be observed in Fig. 13 (e) between the orange and red lines. The maximum pressure drop due to WI or an increase in its quantity ranges between 5 and 10 bar. 5. Analysis of the HRR The observed changes in HRRs with CR, SOI, HES, and WI justify a deep analysis based on the HRR. From Fig. 14(a)–(b), experimental HRRs are depicted for most of the tested conditions. When analyzing their shapes, the HRRs can be categorized into three groups: single mode (SM), single mode with partial low rate (SML), two modes (TM). Due to the complexity of the combustion process and the need for comprehensive explanation, a Computational Fluid Dynamics (CFD) approach has been employed to simulate the three observed combustion modes while maintaining the same conditions as those measured in the tests. The simulation was conducted using Ansys-FORTE 2022 R2, specialized software designed for internal combustion engines. The combustion process has been solved by applying axial symmetry to tested conditions, and the combustion chamber, with the same geometry as the piston for CR 17.5:1, has been divided into 8 sectors. Fig. 15 shows the temperature field output for three of the tested conditions, representative of the HRRs shapes, all of them at CR 17.5:1 and 2000 rpm, where the temperature scale is different for every picture. The presented crankshaft angles are −5◦, −2◦, 0◦, +5◦, +10◦, and +20◦. Since H 2 is supplied through the intake port along with the air, it is assumed that the fuel has had sufficient time (~360◦referred to the crankshaft angle) to thoroughly mix with the air, resulting in an almost homogeneous fuel–air equivalence ratio throughout the combustion chamber. From the experimental HRRs, the combustion process is explained: •SM, where the HRR is fully dominated by a single pulse associated with a premixed combustion process. In this case, once SOC is achieved, the energy of the fuel is released progressively, burning in a single step in a continuous fashion. These are the cases shown in Fig. 14(a) with HES 0 % (only diesel fuel to maintain RPM), Fig. 14 (b) with HES 77 % (magenta) and 80 % (orange), both with WI, and Fig. 14(c) with HES 80 % and WI. In this mode, there is only one relative maximum of the HRRs, and the shape is almost symmetrical around a vertical line that passes through the maximum, that increases almost linearly with HES, irrespective of the presence of Fig. 13. (continued). J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 22 Fig. 14. (a) HRR at CR 17.5, 2000 rpm and SOI 10 ◦with different HES and H 2 O flow (b) HRR at CR 16.3, 2000 rpm and SOI 10◦with different HES and H 2 O flow, (c) HRR at CR 15.5, 2000 rpm and SOI 10◦with different HES and H 2 O flow, (d) HRR at CR 14.5, 2000 rpm and SOI 10◦with different HES and H 2 O flow, (e) HRR at CR 13.5, 2000 rpm and SOI 10◦with different HES and H 2 O flow. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 23 water. This combustion mode occurs at very low loads (only pilot injection) and at the highest loads supported by H2, regardless WI and CR. •SML. In this case, combustion appears to partially resemble a premixed mode, with only one relative maximum in the HRR, but a substantial portion of the fuel is burned at low speeds, extending the combustion duration. These cases are illustrated in Fig. 14(a)–(e) for HES 45 % (blue) without WI and for HES 62 % with and without WI (red). The shape of the HRR has deviated from being symmetrical, and in terms of liberated energy, this segment of the HRR accounts for approximately 50 %. Therefore, it is likely to contribute to a reduction in combustion efficiency and potentially an increase in the amount of unburnt H2. Nonetheless, the maximum HRR increases with HES in an almost linear fashion. This combustion mode primarily occurs at medium loads, within a range of 45 %–77 % HES, although it is influenced by CR. Similar to the other combustion mode, WI has no impact on the shape of the HRR and only causes a delay in the SOC. •TM, where HRRs exhibit two relative maxima that define a dual combustion mode. This occurs in cases where the HES is above ~ 70 %, typically at medium–high loads, as can be observed in Fig. 14(a)– (e). For medium loads, the first HRR maximum is greater than the second one, but for higher HES values, the second maximum exceeds the first, and the energy liberated under this second combustion mode is also higher. This combustion mode primarily occurs at lower compression ratios, ranging from 15.5:1 and below. It also manifests at medium to high loads, specifically within the range of 62 % to 80 % HES. WI alters HRR by delaying the SOC, but it does not significantly change the HRR’s shape. From the temperature field provided by the simulation in Fig. 15, the combustion processes are compared: The first point to consider is the different SOC. At −5◦, combustion has not started in any mode, but at −2◦, the temperature field has changed significantly. The SML exhibits the highest temperature, indicating that the SOC is at its minimum, as shown in Fig. 14(a). At this Fig. 14. (continued). J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 24 angle, the effect of WI is also evident. Despite the higher amount of H 2 in the TM, the temperature is more uniform within the combustion kernel. Between −5◦and +10◦, the SM reveals that the combustion kernel is concentrated around the bowl, justifying the presence of only one peak in the HRR. Conversely, in the cases of SML and TM, two distinct hot areas are clearly visible: one in the bowl and, after several crankshaft angles, another outside the bowl but close to the head of the combustion chamber, attributable to the higher H 2 concentration. Although temperatures are higher in the SML compared to the TM, the difference in maximum temperature between the two hottest areas is smaller in the Fig. 15. Temperature evolution at 2000 rpm and SOI 10 ◦at −5◦, −2◦, 0◦, +5◦, +10◦, +20◦. A) HES 0 % SM, B) HES 62 % SML, C) HES 77 % with WI, TM. J.M. Rueda-V´ azquez et al. Applied Thermal Engineering 238 (2024) 121933 25 SML than in the TM, which explains the second peak in the HRR. Regarding NOx emissions, Fig. 5(a) shows that TM has low NOx at 500 ppm, compared to the 900 ppm of SML, despite having a higher amount of H 2 . In Fig. 15, it can be observed that the maximum temperature in TM is consistently lower than in SML at any crank angle, which can justify the lower level of NOx in TM. 6. Conclusions This work presents an experimental investigation of the effects of using H 2 as the primary fuel in an ICE. The study explores NOx emissions, efficiency, MEP, SOC, and maximum in-cylinder pressure through experiments, comparing them with WI and CR in a modified high-speed turbocharged common-rail automotive diesel engine. Based on the experimental results, the following conclusions can be drawn from this investigation: •NOx emissions increase monotonically with CR for any HES, with the rate of increase accelerating as HES and RPM decrease. Within the CR range of 13.5 to 17.5 at 2000 RPM, NOx increases by 80 % at HES 62 %, but at HES 45 %, NOx increases by 400 %. •NOx emissions increase monotonically with HES for any CR, RPM, or SOI. At 2500 rpm, SOI 10◦, CR 17.5, and without WI, going from HES 38 %–55 % (a 45 % increment) results in a 57 % increase in NOx. However, going from HES 55 %–64 % (a 16.3 % increment) leads to a 54 % increase in NOx. This behavior demonstrates the non-linear response of NOx to changes in HES, with the increase in NOx relative to the HES increase being almost cubic. •Injection of water reduces NOx for the same HES, but NOx tends to be more sensitive to changes in HES than to changes in WI. At 2000 rpm, SOI 10◦, and CR 17.5, NOx decreases from 900 ppm to 370 ppm when WI increases from 0 to 0.266 g/cycle. For 77 % HES and CR 16.3, NOx decreases by 41 % with a 100 % increase in WI. Therefore, the effect of WI on NOx reduction becomes less significant as HES increases; as HES increases, the required WI should also increase, following an almost quadratic pattern. •NOx decreases monotonically with RPM, significantly reducing the range of NOx levels with HES and diminishing the influence of CR and WI on it. For SOI 10◦at 2000 rpm, the NOx range is between ~20 and 1000 ppm for all engine loads and WI levels. With the same SOI at 2750 rpm, the NOx levels range from ~20 to 590 ppm. Consequently, an increase in RPM of 37.5 % reduces the maximum NOx by 40 %, resulting in a ratio of ΔRPM/ΔNOx of approximately −1 for this SOI. •SOI has a significant impact on NOx emissions, as even a small reduction in SOI results in a substantial decrease in NOx. For all HES, WI, and CR settings at 2000 rpm with SOI at 10◦, the NOx range is between ~20 and 1000 ppm. When the SOI is reduced to 0◦at the same RPM, the NOx range narrows to ~20 and 350 ppm. Therefore, a 50 % reduction in SOI leads to a 65 % reduction in NOx, resulting in a ΔSOI/ΔNOx ratio of approximately 0.8. •The combination of different agents, including WI, SOI, CR reduction, and RPM increase, all managed simultaneously, exhibits an accumulative pattern. However, it’s important to note that the effects of these different agents cannot be simply added together as if they were acting independently. For example, with an HES of 72 % at SOI 10◦, 2000 rpm, CR 17.5, and WI of 0.266 g/cycle, the NOx level is 650 ppm. In contrast, when keeping the same HES but reducing SOI to 5◦, increasing RPM to 2750, raising WI to 0.388 g/cycle, and decreasing CR to 14.5, it allows for a reduction in NOx to 100 ppm. •The parameter α has proven to be highly effective in determining the maximum amount of WI needed to control NOx emissions, ensuring it meets the required limits. As a result, this parameter can be utilized in designing the strategy for NOx control. •It has been observed that reducing the CR from 17.5:1 to 14.5:1 has a negligible effect on efficiency, remaining nearly constant. It is only when the CR is reduced to 13.5:1 that a decrease in efficiency, approximately 5 %, is observed, which can be attributed to the efficiency cycle penalty. Therefore, it is feasible to maintain high efficiency up to CR 14.5:1 while achieving low NOx levels. •The operation of the engine using H 2 as the primary fuel is stable and knock-free when combined with WI, up to a MEP of approximately 11.5 bar, the highest value achieved. •The in-cylinder peak pressure decreased significantly with the reduction in CR. The maximum pressure drop observed was 40 bar between CR 17.5:1 and 13.5:1 for the highest HES. •WI has several effects regardless of CR: (i) up to a 5 % reduction in efficiency, (ii) a maximum MEP loss of 0.5 bar compared to cases without WI or with a lower amount of water, (iii) a SOC by one or two crank angle degrees, and (iv) a maximum in-cylinder pressure decreases of between 5 and 10 bar. •It has been demonstrated that it is feasible to transform current diesel engine technology to operate in a stable dual-mode, knock-free manner, where diesel is used solely for ignition. This allows for achieving high efficiency while maintaining minimal NOx levels with WI. These results make it possible to consider H 2 as a viable fuel option for current diesel engine technology, serving as an alternative to fuel cells due to its robustness, cost-effectiveness, high efficiency, and the significant levels of MEP achieved. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgements This work was supported by the Junta de Andalucía (Consejería de Economía, Conocimiento, Empresas y Universidad) as part of item PY20 RE 010 within the “Incentivos a los agentes del sistema andaluz del conocimiento – Ayudas a la I +D +i, en el ´ ambito del Plan Andaluz de Investigaci´ on, Desarrollo e Innovaci´ on (PAIDI 2020). Authors are also thankful to the Spanish Ministry of Science and Innovation (PID2019- 105936RB-C21 and TED2021-130596B-C22). Annex. Maximum and minimum values for the H 2 fuel mode studied. SOI for pilot injection H 2 (kg/ h) HES (%) H 2 O (kg/ h) Diesel (kg/ h) Torque (Nm) Power (kW) MEP (bar) Efficiency Fuel/Air equivalence ratio CR 17.5:1 (continued on next page) J.M. Rueda-V´ azquez et al.