Modelling Ammonia-Hydrogen-Air Combustion and Emission Characteristics of a Generic Swirl Burner

Abstract

Abstract Ammonia (NH3) is considered a valued alternative carbon-free fuel able to reduce greenhouse gas (GHG) emissions with respect to conventional fuel, therefore it has attracted the interest of gas turbine manufacturers over the last years. However, the use of ammonia in existing gas turbine combustors entails not yet resolved challenges: based on currently available studies, the simultaneous coexistence of high NOx emissions and poor flame stability is expected when typical lean premixed technologies are used. NOx containment strategies may lead to significant combustion architectural changes. The use of NH3/H2 blends is a valued way to improve the flame stability, even though in some cases it is obtained to the detriment of the emission performances. However, there is still a lack of knowledge and experimental data on the use of ammonia and hydrogen-ammonia blends under typical operating conditions of gas turbines. The combustion process of both pure NH3 and a NH3/H2 fuel blends is here analysed using two kinetics processors, i.e. CHEMKIN-PRO and CANTERA: detailed reaction mechanisms have been tested and compared in terms of laminar flame speed, ignition delay times and extinction strain rate with the aim to identifying the most suitable ones for the evaluation of NOx emissions. The generic swirl burner being used in Cardiff University’s Gas Turbine Research Centre has been considered as validation test case. Evaluations have been carried out by means of 2D axisymmetric RANS simulations leading to a significant reduction of the computational time. Different pressures and mass flow rates are evaluated to understand the correlation in the formation of NOx emissions for pollutants reduction purpose. A direct comparison between experimental and numerical results is carried out in terms of flow field, flame shape and NOx emissions. Results show that NOx emissions decrease significantly with increase in pressure, also indicating guidelines for using a simplified RANS analysis, which leads to improved computational efficiency, allowing wide sensitivity and optimisation analyses to support the design development of an industrial combustion system.