Flame Stabilization and Emission Characteristics of a Prototype Gas Turbine Burner at Atmospheric Conditions Available to Purchase

In the present work, a downscaled prototype 4^th generation Dry Low Emission gas turbine (SGT-750) burner (designed and manufactured by Siemens Industrial Turbomachinery AB, Sweden) was investigated using an atmospheric experimental facility. The primary purpose of the research is to analyze flame stability and emission capability of the burner. OH Planar Laser-Induced Fluorescence (OH-PLIF), and chemiluminescence imaging were performed to characterize the flame structure and location. From the OH-PLIF images, the reaction zone and post flame region could be identified clearly. The chemiluminescence images provide an estimation of the overall heat release from the secondary combustion zone inside the Quarl. Emission was measured using a water-cooled emission probe, placed at the exit of the combustor to sample NOx and CO concentrations. The global equivalence ratio (Φ) was varied from rich to lean limit (flame temperature change from 1950 K to 1570 K) for understanding the stable and instable reaction zones inside the Quarl. Total thermal power was varied from 70 kW to 140 kW by changing global Φ and burner throat velocity (60 to 80 m/s). Near the lean blowout (LBO) event (at global Φ ∼ 0.4), instability of reaction zone is revealed from the flame images. Incorrect modulation of Pilot and RPL fuel splits show instable flame. Flame instability mitigation was possible using higher amount of RPL and Pilot fuel (trade-off with emission performance). The main flame LBO margin was extended by applying higher Pilot fuel and using higher preheat air temperature. Numerical analysis was carried out using Fluent to understand the scalar and vector fields. A basic chemical reactor network model was developed to predict the NOx and CO emission with experimental results. NOx emission prediction showed good agreement with experiment; whereas the model is failed to capture accurate CO emission in the lean operating points.