Abstract
Ammonia is emerging as a very convenient hydrogen- and energy-carrier chemical that, in the context of present efforts to curb carbon emissions from the power-generation and transport sectors, can be manufactured from both green hydrogen, generated from renewable energy sources, and blue hydrogen, generated from natural gas in combination with carbon capture and storage. As opposed to hydrogen, however, the properties of ammonia make it significantly simpler to transport across long distances and store for relatively long periods of time. Early exploratory work on combustion of pure ammonia in laboratory-scale gas turbine burners was conducted at Tohoku University and AIST (Japan) and revealed that the adoption of a longitudinal rich-lean staging strategy in the operation of the device is a convenient approach to minimize NOx and N2O emissions from fuel-bound nitrogen oxidation. Moreover, very recent experimental evidence acquired at SINTEF using a laboratory-scale 4th-generation DLE burner design by Siemens Energy (SGT-750) confirms that the low-emission performance achieved with rich-lean staging also applies to combustion of partially-decomposed ammonia [ASME paper GT2021-60057]. In the present paper we report a comprehensive numerical modelling study that exploits Large Eddy Simulation (LES) in conjunction with detailed chemical kinetics and a Chemical Reactor Network (CRN) model to assess a rich-lean staging strategy applied to combustion of partially-decomposed ammonia in the Siemens Energy 4th-generation DLE burner. Data analysis performed from both numerical modelling approaches, LES and CRN, confirm that the rich-lean staging strategy tested in the present study indeed results in significantly lower emissions compared to the conventional operational profile of the burner. Furthermore, reaction pathways analysis performed on the CRN data reveals important details that characterize the different evolution of nitrogen species between the non-staged and staged operation of the burner, ultimately leading to the observed difference in NOx and N2O emissions.
