This work is a numerical study of the use of ammonia and hydrogen in a high-pressure-dual-fuel (HPDF) combustion. The main fuels (hydrogen and ammonia) are direct injected and ignited by a small amount of direct injected pilot fuel. The fuels are injected using a dual fuel injector from Woodward L’Orange, which can induce two fuels independently at high pressures up to 1800 bar for the pilot fuel and maximum 500 bar for the main. The numerical CFD-model gets validated for of hydrogen-HPDF with experimental data. Due to safety issues at the test rig it was not possible to use ammonia in the experiments, so it is modelled using the numerical model. It is assumed that the CFD-model also gives qualitative correct results for the use of ammonia as main fuel, so a parameter study of ammonia-HPDF is made. The results for the hydrogen-HPDF show, that hydrogen can be used in the engine without any further modifications. The combustion is very stable, and the hydrogen ignites almost immediately when it enters the combustion chamber. The results of the ammonia combustion indicate, that the HPDF combustion mode can handle ammonia effectively. It seems beneficial to inject the ammonia at higher pressures than hydrogen. Also pre-heating the ammonia can increase the combustion efficiency.

Further improvement of hydrogen combustion in port fuel injection engines is limited by backfire. To overcome this drawback of hydrogen port fuel injection engines it is essential to locate and understand the reasons for the inflammation of a backfiring cycle. To contribute to this understanding a minimal invasive lateral optical access was developed for a medium speed large bore engine. The access uses a UV enhanced endoscope to investigate the OH radical’s natural chemiluminescence to locate the inflammation of a backfiring cycle in the combustion chamber. The investigations are carried out at high engine load. The optical investigations were based on a thermodynamic screening. This included the variation of the start of the hydrogen port fuel injection and the engine’s backpressure. These experiments prove the influence of exhaust backpressure and the start of injection on the probability of backfire. As higher backpressure leads to an increased probability of backfire, the SoI strategy has also a decisive influence. An optimum start of injection timing with less backfire under high backpressure was experimentally determined at 300°CA with respect to 720°CA as FTDC. The conducted optical investigations show that backfire starts by ignition by hot residual gasses during the first cycle located under the exhaust valves. Furthermore, the results show ongoing combustion in the intake manifold leading to serious damage of the engine if not prohibited.

The following paper presents thermodynamic and optical investigations of the natural flame and OH radical chemiluminescence of a hydrogen enriched methane combustion compared to natural gas combustion. The engine under investigation is a port-fueled unscavenged prechamber 4.8 L single cylinder large bore engine. The blends under consideration are 2%_V, 5%_V,10%_V, and 40%_V of hydrogen expected to be blended within existing natural gas grids in a short and mid-term timeline in order to store green energy from solar and wind. These fuel blends could be used for stabilization of the energy supply by reconverting the renewable fuel CH₄/H₂ in combined heat and power plants. As expected, admixture of hydrogen extends the ignition limits of the fuel mixture toward lean ranges up to an air-fuel equivalence ratio of almost 2. No negative effect on combustion is observed up to an admixture of 40%_V hydrogen. At 40%_V hydrogen, abnormal combustion like backfire occurs at an air-fuel equivalence ratio of 1.5. The higher mixtures exhibit increased nitrogen oxide emissions due to higher combustion chamber temperatures, while methane slip and CO emissions are reduced due to more complete combustion. The optical investigation of the natural flame and OH radical chemiluminescence are in good agreement with the thermodynamic results verifying the more intense combustion of the fuel blends by means of the chemiluminescence intensity. Further, lube oil combustion and a continuing luminescence after the thermodynamic end of combustion are observed.

High load (18 bar IMEP) dual fuel combustion of a premixed natural gas/air charge ignited by directly injected diesel fuel was studied in a large bore gas engine. A nozzle design with low flow rate was installed to inject a small diesel volume (10.4 mm3) equal an energetic amount of about two percent. The effect of compression end temperature on ignition and combustion was investigated using valve timings with early IVC (Miller) and maximum charging efficiency (MaxCC). Furthermore, the engine speed was reduced (1500 rpm to 1000 rpm) for the Miller valve timing to analyze the impact of the chemical time scale on the combustion process. During all experiments, the cylinder charge density was kept constant adjusting the intake pressure and the resulting air mass flow. Unlike a typical reactivity-controlled compression ignition (RCCI) combustion process, the combustion phasing of the investigated pilot-ignited natural gas combustion was also sensitive to the injection timing besides the air-fuel equivalence ratio (AFER). As a consequence of the lower compression end temperature with Miller valve timing, the operating range of the AFER was shifted to significantly lower values (λ = 1.53 to 1.63) compared to MaxCC (λ = 1.88 to 2.08) and knocking combustion was no longer observed. Furthermore, the bijective relation between the mass fraction burnt 50% (MFB50%) as a function of the start of energizing (SoE) at a constant air-fuel equivalence ratio was no longer valid. The experiments with MaxCC valve timing revealed that this behavior is not necessarily a characteristic of the pilot ignition. In fact, conditions inhibiting the ignition and engine knock result in loss of the one-to-one pairing between SoE and MFB50%. The reduction of the engine speed extended the lean misfire limit (λ = 1.53 to 1.75) and improved the engine start behavior.

Emissions of nitrogen oxides (NOx) from marine propulsion systems have gained public interest resulting in emission limits as defined by the International Maritime Organization (IMO) with IMO Tier III, especially for vessels operating in Emission Control Areas (ECA). The reduction of greenhouse gas emissions is also increasingly important for marine propulsion. Minimizing NOx while reducing climate impact calls for technologies such as the gas engine with aftertreatment systems, preferably with the ability to run on alternative fuels. A proven technology for reducing NOx in marine engines is the Selective Catalytic Reduction (SCR) aftertreatment system. It is also possible to avoid engine raw emissions by shifting the combustion process to lower temperature levels. Hydrogen is an alternative fuel with combustion properties enabling premixed operation at significantly higher air-fuel ratio than natural gas (NG) and thus, reducing raw NOx emissions. The study uses a systematic approach to compare emissions and efficiency of a lean-burn gas engine with a natural gas and a mild conversion hydrogen setup, utilizing two different strategies: combustion of NG with the assumption of an SCR catalyst and high raw NOx emissions and combustion of pure hydrogen using the NOx reduction potential of higher excess air. The scope of the study makes it possible to illustrate engine concepts for future applications in the displacement class of 4.8 L per cylinder. The highest efficiency of 45.3% was achieved with the natural gas engine and SCR. The concept with the lowest Global Warming Potential (GWP) was the hydrogen fueled engine under the prerequisite of using green hydrogen, accompanied by a reduction in efficiency of 0.6% compared to the efficiency optimum of NG with SCR. Assuming the use of gray hydrogen, the GWP was 48% and 52% higher than with NG and NG with SCR, respectively, at the efficiency-optimal operating points.