| description abstract | Humidified gas turbines (HGT) offer the attractive possibility of increasing plant efficiency without the cost of an additional steam turbine as is the case for a combined gassteam cycle. In addition to efficiency gains, adding steam into the combustion process reduces NOx emissions. It increases the specific heat capacity (hence, lowering possible temperature peaks) and reduces the oxygen concentration. Despite the thermophysical effects, steam alters the kinetics and, thus, reduces NOx formation significantly. In addition, it allows operation using a variety of fuels, including hydrogen and hydrogenrich fuels. Therefore, ultrawet gas turbine operation is an attractive solution for industrial applications. The major modification compared to current gas turbines lies in the design of the combustion chamber, which should accommodate a large amount of steam without losing in stability. In the current study, the premixed combustion of pure hydrogen diluted with different steam levels is investigated. The effect of steam on the combustion process is addressed using detailed chemistry. In order to identify an adequate oxidation mechanism, several candidates are identified and compared. The respective performances are assessed based on laminar premixed flame calculations under dry and wet conditions, for which experimentally determined flame speeds are available. Further insight is gained by observing the effect of steam on the flame structure, in particular HO2 and OH* profiles. Moreover, the mechanism is used for the simulation of a turbulent flame in a generic swirl burner fed with hydrogen and humidified air. Large eddy simulations (LES) are employed. It is shown that by adding steam, the heat release peak spreads. At high steam content, the flame front is thicker and the flame extends further downstream. The dynamics of the oxidation layer under dry and wet conditions is captured; thus, an accurate prediction of the velocity field, flame shape, and position is achieved. The latter is compared with experimental data (PIV and OH* chemiluminescence). The reacting simulations were conducted under atmospheric conditions. The steamair ratio was varied from 0% to 50%. | |
