A Study on Combustion Instability Analysis of Partially Premixed Model Gas Turbine Combustor with 1D Lumped Method

Abstract

Recently, fine dust generation in the East Asian region has been emerging as a serious problem in each spring. It is expected that the strategy of increasing the weight of the combined cycle power plant, which is an environmentally friendly type of power generation, will play a big role in reducing fine dust. Gas turbines that use various renewable fuel such as syngas, synthetic natural gas (SNG), and biogas are being steadily developed, as it is effective not only to reduce fine dust but also to reduce exhaust emissions. Also, partially premixed gas turbine combustor is basically designed as a lean burn and has the possibility of manifesting combustion instability phenomenon. Therefore, it is important to understand and predict combustion instability characteristics for various fuel compositions in gas turbine combustor development. The purpose of this study is to analyze the instability characteristic of partially premixed combustor according to combustor length, flame position and fuel composition using 1D lumped method. In addition, the prediction of the instability mode shifting, which is the unique phenomenon of partially premixed combustor, is conducted for the first time. As a study procedure, Flame Transfer Function (FTF) in various fuel compositions are obtained with photomultiplier tube (PMT) and hot wire anemometry (HWA). Then the length of the combustor and the position of the flame obtained through the OH-PLIF image were adjusted. In addition, the thermal properties for various fuel compositions were applied using the NASAs CEA code. The predictions of instability frequency of dominant longitudinal mode and multimode instabilities were similar to those of experimental results. As the length of the combustor increases, a decrease in the combustion instability frequency and a change in the instability mode are were predicted. Also, by analyzing the combustion instability characteristics according to the flame position, it is confirmed that the instability mode shifting phenomenon can be predicted only by the change of the flame position under the same FTF and fuel/air condition. In the overall fuel composition, the instability mode was predicted to increase as the percentages of H2 fuel composition increased, as in the combustion instability mode shifting phenomenon that occurred in the combustion experiments. Prediction of instability frequency and instability amplitude trends were improved by applying combustor temperature and reflection coefficient of experiment. Based on these results, it is possible to predict the instability amplitude tendency and suggest the direction to avoid instability during design and operation.