Combustion Instability Diagnosis Using Machine Learning Methodology of High-Speed Flame Images for the Safe Operation of a Gas Turbine Combustor

2021 ◽  
Vol 45 (8) ◽  
pp. 447-458
Author(s):  
Byung Chul Jun ◽  
Dae Jin Jang ◽  
Min Chul Lee
Author(s):  
Jisu Yoon ◽  
Seongpil Joo ◽  
Min Chul Lee ◽  
Jeongjin Kim ◽  
Jaeyo Oh ◽  
...  

Recently, energy resource depletion and unstable energy prices have become global issues. Worldwide pressure to secure and make more gas and oil available to support global power needs has increased. To meet these needs, alternative fuels composed of various types of fuels have received attention, including biomass, dimethyl ether (DME), and low rank coal. For this reason, the fuel flexibility of the combustion system becomes more important. In this study, H2 and CH4 were selected as the main fuel composition variables and the OH-chemiluminescence measurement technique was also applied. This experimental study was conducted under equivalence ratio and fuel composition variations with a model gas turbine combustor to examine the relation between combustion instability and fuel composition. The combustion instability peak occurs in the H2/CH4 50:50 composed fuel and the combustion instability frequency shifted to higher harmonic of longitudinal mode based on the H2 concentration of the fuel. Based on instability mode and flame length calculation, the effect of the convection time during the instability frequency increasing phenomenon was found in a partially premixed gas turbine combustor. The time-lag analysis showed that the short convection time in a high H2 concentration fuel affects the feedback loop period reduction and, in these conditions, high harmonics of longitudinal mode instability occurs. This fundamental study on combustion instability frequency shifting characteristics was conducted for H2/CH4 composed fuel and the results contribute key information for the conceptual design of a fuel flexible gas turbine and its optimum operation conditions.


2011 ◽  
Vol 21 (1) ◽  
pp. 013124 ◽  
Author(s):  
Hiroshi Gotoda ◽  
Hiroyuki Nikimoto ◽  
Takaya Miyano ◽  
Shigeru Tachibana

Author(s):  
Xiaoling Chen ◽  
Wyatt Culler ◽  
Stephen Peluso ◽  
Domenic Santavicca ◽  
Jacqueline O’Connor

Low-emissions gas turbine combustion, achieved through the use of lean, premixed fueling strategies, is susceptible to combustion instability. The driving mechanism for this instability arises from fluctuations of pressure, fuel/air flow rate, and heat release rate. If these fluctuations are relatively in-phase, the combustion system will evolve to a self-excited state. The self-excited instability frequency and amplitude depend mainly on the operating condition and the geometry of the combustor. In this study, we consider the onset and decay of self-excited instabilities, resulting from transients in fuel/air ratio, in both single-nozzle and multi-nozzle combustors. In particular, we examine the differences in the instability onset and decay processes between these two flame configurations, as most gas turbine combustors have multiple nozzles, but most gas turbine combustor experiments utilize a single-nozzle. A nonlinear logistic regression analysis is applied to study the timescales of the decay and onset transients. Variations in the equivalence ratio change the heat release rate distribution inside the combustor, which is captured using chemiluminescence imaging. The normalized Rayleigh index, which shows the spatial distribution of the instability driving, is calculated to analyze the driving strength in different regions of the flame. Comparisons between the single- and multi-nozzle flame transients, including both center and outer flames for the multi-nozzle combustor, suggest that both confinement from the wall and flame-flame interaction are crucial to determining flame dynamics as the equivalence ratio transient changes the heat release rate distribution near corner recirculation zone and flame shear layers.


Sign in / Sign up

Export Citation Format

Share Document