Numerical Investigation of Lean Neat Ammonia Combustion in a Heavy-Duty Diesel Engine Converted to Spark Ignition

Ammonia (NH₃) use as fuel poses technical challenges such as increased nitrogen-based and unburned NH₃ emissions. This study used a 0D model coupled with detailed NH₃ kinetics to evaluate the effect of equivalence ratio (ϕ) from 0.7 to 1.0 in a heavy-duty compression ignition engine converted to spark ignition operation. The goal was to evaluate how ϕ affected NO_x and N₂O formation and/or destruction at constant fuel energy per cycle, engine speed, and CA50. Simulated NO_x emissions (i.e., NO + NO₂) followed a trend similar to the one typically observed for hydrocarbon fuels in a SI engine, but that was different from the experiment. In addition, it underpredicted NO_x emissions for ϕ = 0.7 by 79% and overpredicted NO_x emissions for ϕ = 1 by 576%. The simulation showed that thermal NO production was more than 80% from the total NO production, but the effect of ϕ on this percentage was negligible. Then, predicted N₂O emissions had an opposite trend and were three orders of magnitude lower than the experiment. Under the assumption that post-combustion phenomena can explain these differences, an additional reactor simulating of the chemistry inside the unburned mixture exiting crevices post combustion at ϕ = 0.9 predicted N₂O production similar to experimental data and an additional 20 ppm of NO. A third reactor simulating the exhaust blowdown predicted substantial DeNO_x reactions (up to 3000 ppm decrease in NO_x), and a large N₂O production above experimental values. Therefore, the significant DeNO_x and N₂O formation reactions could explain the differences between the 0-D engine simulation and experiments even when accounting for the real mixture inhomogeneities, which emphasizes the importance of capturing both in-cylinder and post-EVO conditions when modeling NO_x and N₂O emissions in an IC engine.