The maritime sector is under increasing pressure to reduce greenhouse gas (GHG) emissions, with the International Maritime Organization (IMO) aiming for net-zero GHG shipping by 2050. Replacing oil-fueled diesel engines with alternative fuel engines is crucial. Ammonia, hydrogen, methanol, and methane are potential alternatives, categorized as gray/brown (using fossil fuels without carbon capture), blue (fossil fuels with carbon capture), or green (renewable energy sources). The IMO emphasizes avoiding emissions shifting to other sectors, thus favoring blue/green fuels. For absolute-zero emissions, only carbon-free fuels like hydrogen and ammonia are viable. Ammonia offers advantages of low storage and transportation costs and existing infrastructure. Studies have shown it to be a cost-effective and environmentally friendly alternative compared to methanol and methane, producing no CO2, particulate matter, CO, hydrocarbons, or formaldehyde during combustion. The main challenge is ammonia's poor combustion characteristics, such as high ignition energy, a narrow flammability limit, and slow propagation speed. Pilot-diesel-ignition combustion modes (high-pressure dual-fuel (HPDF) and low-pressure dual-fuel (LPDF)) have been explored, but both suffer from high unburned ammonia and N2O emissions. Current strategies like split diesel injection and early diesel injection aim to reduce unburned ammonia but often lead to increased NOx emissions, creating a trade-off. This necessitates innovative technologies to overcome these limitations and achieve efficient and clean ammonia combustion.

Existing research highlights the challenges and opportunities of using ammonia as a marine fuel. Studies comparing different alternative fuels (e.g., hydrogen, ammonia, methanol, methane) have shown ammonia's potential as a balanced carbon-free option. The low storage, transportation costs, and established infrastructure make it attractive for international shipping. Lifecycle cost analyses also favor ammonia over hydrogen and methanol. While ammonia combustion is inherently cleaner than carbonaceous fuels, the poor combustion characteristics present a significant hurdle. High-pressure and low-pressure dual-fuel combustion strategies have been investigated, but issues such as high unburned ammonia and N2O emissions, as well as a trade-off between unburned ammonia and NOx emissions remain. Efforts to improve these modes through techniques like split diesel injection and early injection have shown limited success, often resulting in increased NOx or N2O. The need for new approaches to solve these challenges is evident.

This study proposes a novel in-cylinder reforming gas recirculation (IRGR) concept to address the limitations of current ammonia combustion technologies. A four-cylinder pilot-diesel-ignition ammonia dual-fuel engine without IRGR served as the baseline for experimental data acquisition and model validation. A chemical reaction mechanism for combined ammonia and n-heptane combustion (65 species and 344 reactions) was developed and integrated into a numerical simulation platform (CONVERGE). The numerical model was validated against experimental data from the base engine under varying ammonia energetic ratios (40%, 60%, 80%, 90%), engine loads (50%, 75%, 85%, 100%), and injection timings. The IRGR concept involves one cylinder operating rich in ammonia, partially decomposing it into hydrogen under high-temperature conditions. The exhaust from this 'reforming cylinder' (containing H2, NH3, CO, O2, N2, CO2, CH4, NO, NO2, and N2O) is recirculated into the other cylinders. This combines the advantages of hydrogen enrichment and exhaust gas recirculation (EGR). The numerical simulations investigated the IRGR concept's effects on thermal efficiency, unburned NH3, NOx, N2O, and GHG emissions under different operating conditions (ammonia energetic ratios: 80%, 90%, 97%; engine speeds: 1000, 1500, 2000, 2500 rpm; injection timings: -20° to 0°CA aTDC; and reforming cylinder enrichment levels). The simulations used a detailed 3D computational fluid dynamics (CFD) approach with adaptive mesh refinement to capture the complex flow and combustion processes. The Lagrangian particle method modeled diesel fuel injection, including spray breakup, droplet drag, collision, and evaporation models. The SAGE solver simulated combustion. The study compared the IRGR engine's performance and emissions to those of the base engine to evaluate the effectiveness of the proposed concept.

The simulations accurately predicted heat release, in-cylinder pressure, and exhaust emissions of the base engine. In the dedicated reforming cylinder, hydrogen generation was highest near the high-temperature regions where ammonia decomposition and reforming reactions were most favorable. At 0.7 and 0.8 overall excess air ratios, hydrogen conversion ratios were 32.0% and 20.8%, respectively. NO and N2O formation in the reforming cylinder was minimal due to oxygen deficiency. In the hydrogen-enriched cylinders, hydrogen addition improved combustion, reducing unburned ammonia in the clearance volume and suppressing NO and N2O formation due to the reduced oxygen content. At 1000 rpm and 10% diesel energetic ratio, IRGR reduced unburned ammonia by 84%, N2O significantly, and NOx considerably. Unburned losses decreased from 11.6% to 3.0%, improving fuel economy. Under various ammonia energetic ratios (80%, 90%, 97%) and injection timings, IRGR consistently reduced NOx and unburned NH3 and increased indicated thermal efficiency. Even at 97% ammonia energetic ratio, IRGR reduced unburned ammonia to about 8 g/kWh and increased efficiency to around 46%. At higher engine speeds (1500, 2000, 2500 rpm), IRGR maintained its effectiveness, significantly improving thermal efficiency and emission reductions, particularly important at 2500 rpm where N2O emissions from the base engine reached levels comparable to pure diesel. At 3% diesel energetic ratio and 1000 rpm, IRGR increased indicated thermal efficiency by 15.8%, reduced unburned NH3 by 89.3%, N2O by 91.2%, and GHG emissions by 83.7% compared to the base engine. Compared to pure diesel, GHG reduction reached 94%. At 20% diesel and 2500 rpm, IRGR improved thermal efficiency by 24.4% and reduced NH3 by 86.6% and N2O by 83.7%.

The IRGR concept effectively addresses the key challenges associated with ammonia combustion in marine engines. The simultaneous hydrogen enrichment and EGR significantly improve combustion efficiency, reducing unburned ammonia and N2O. The reduction in oxygen content from EGR suppresses NOx formation. The results demonstrate the potential of IRGR to achieve significant GHG emission reductions while improving fuel economy, overcoming the trade-off often seen between unburned ammonia and NOx. The findings provide a strong foundation for developing cleaner and more efficient ammonia-fueled marine engines, aligning with the IMO's decarbonization targets. The high level of emission reduction, especially at high engine speeds, highlights the potential for substantial environmental impact. The use of a detailed chemical kinetic mechanism and 3D-CFD modeling ensures the robustness of the findings.

This study successfully demonstrated the potential of in-cylinder reforming gas recirculation (IRGR) to significantly improve the efficiency and reduce emissions of pilot-diesel-ignition ammonia combustion engines. IRGR achieves substantial reductions in unburned NH3, N2O, and greenhouse gases while significantly increasing thermal efficiency, even at high engine speeds and high ammonia energetic ratios. Future work should focus on experimental validation of the IRGR concept and optimization of the system parameters for large-scale marine engine applications. Further research could explore different reformer designs and integration strategies to enhance the system's overall performance and durability.

The study is primarily based on numerical simulations. While the numerical model was validated against experimental data from a base engine, experimental validation of the IRGR concept is needed to confirm the predicted improvements. The computational cost limited the optimization of the pilot diesel injection timing for all cylinders simultaneously. Future research should address these limitations through experimental validation and comprehensive system optimization.