An Experimental Study on the Coupling of Combustion Instability Mechanisms in a Lean Premixed Gas Turbine Combustor

Author(s):  
Hyung Ju Lee ◽  
Kyu Tae Kim ◽  
Jong Guen Lee ◽  
Bryan D. Quay ◽  
Domenic A. Santavicca

An experimental study was conducted to characterize the combined effects of flame-vortex interactions and equivalence ratio fluctuations on self-excited combustion instabilities in a swirl-stabilized lean premixed gas turbine combustor. The combustor was designed so that the fuel injector location and the combustion chamber length could be independently varied. In addition, the fuel and air could be mixed upstream of the choked inlet to the combustor, thereby eliminating the possibility of equivalence ratio fluctuations. Experiments were performed over a broad range of operating conditions and at each condition both the combustor length and the fuel injection location were varied. Dynamic pressure in the combustor, acoustic pressure and velocity in the mixing section, and the overall rate of heat release were simultaneously measured at all operating conditions. Two distinct instability regimes were observed; one near 220 Hz and the other near 345 Hz. It was also found that the strength of the instability changed significantly as the fuel injection location was varied, while the phase of the acoustic pressure and velocity fluctuations in the mixing section did not change. A time series of pressure and CH* chemiluminescence signals confirmed constructive or destructive coupling of the two instability mechanisms; the flame-vortex interaction and the equivalence ratio fluctuation interact each other and determine the instability characteristics in partially premixed conditions.

Author(s):  
Hyung Ju Lee ◽  
Kyu Tae Kim ◽  
Jong Guen Lee ◽  
Bryan D. Quay ◽  
Domenic A. Santavicca

An experimental study was conducted to estimate and confirm equivalence ratio fluctuations at the inlet of a lean premixed gas turbine combustor. Fuel injectors were placed at several locations in the mixing section of the combustor, in order to produce different instability characteristics due to the equivalence ratio fluctuations. An IR absorption technique was used to measure the equivalence ratio fluctuations at the inlet of the dump combustor. The measured IR signals were processed in two different ways and the results were compared to confirm the two calibrated equivalence ratio signals. The processed data showed that the two processing methods gave very similar results, and the phase of the measured equivalence ratio fluctuations at the combustor inlet by the IR absorption technique agreed well with that of equivalence ratio fluctuations predicted by time lags in the mixing section. It was, however, not possible to accurately predict the magnitude of the equivalence ratio fluctuations at the combustor inlet by the time lag analysis because the equivalence ratio fluctuations generataed at the fuel injection location is changed by mixing and diffusion as the fuel is convected through the combustor.


Author(s):  
K. O. Smith ◽  
L. H. Cowell

Rig testing of a lean-premixed, liquid-fueled corabustor was conducted to establish the feasibility of achieving ultra-low NOx emissions at typical gas turbine operating conditions. Two different filming fuel injector concepts were evaluated. The majority of combustor testing was conducted using No. 2 diesel. The test results showed 12 and 20 ppm NOx at 6 and 9 atm, respectively. Corresponding CO levels were 50 ppm in both cases.


Author(s):  
K. O. Smith ◽  
A. Fahme

Three subscale, cylindrical combustors were rig tested on natural gas at typical industrial gas turbine operating conditions. The intent of the testing was to determine the effect of combustor liner cooling on NOx and CO emissions. In order of decreasing liner cooling, a metal louvre-cooled combustor, a metal effusion-cooled combustor, and a backside-cooled ceramic (CFCC) combustor were evaluated. The three combustors were tested using the same lean-premixed fuel injector. Testing showed that reduced liner cooling produced lower CO emissions as reaction quenching near the liner wall was reduced. A reduction in CO emissions allows a reoptimization of the combustor air flow distribution to yield lower NOx emissions.


2005 ◽  
Vol 127 (2) ◽  
pp. 286-294 ◽  
Author(s):  
K. D. Brundish ◽  
M. N. Miller ◽  
C. W. Wilson ◽  
M. Jefferies ◽  
M. Hilton ◽  
...  

The objective of the work described in this paper was to identify a method of making measurements of the smoke particle size distribution within the sector of a gas turbine combustor, using a scanning mobility particle sizing (SMPS) analyzer. As well as gaining a better understanding of the combustion process, the principal reasons for gathering these data was so that they could be used as validation for computational fluid dynamic and chemical kinetic models. Smoke mass and gaseous emission measurements were also made simultaneously. A “water cooled,” gas sampling probe was utilized to perform the measurements at realistic operating conditions within a generic gas turbine combustor sector. Such measurements had not been previously performed and consequently initial work was undertaken to gain confidence in the experimental configuration. During this investigation, a limited amount of data were acquired from three axial planes within the combustor. The total number of test points measured were 45. Plots of the data are presented in two-dimensional contour format at specific axial locations in addition to axial plots to show trends from the primary zone to the exit of the combustor. Contour plots of smoke particle size show that regions of high smoke number concentration once formed in zones close to the fuel injector persist in a similar spatial location further downstream. Axial trends indicate that the average smoke particle size and number concentration diminishes as a function of distance from the fuel injector. From a technical perspective, the analytical techniques used proved to be robust. As expected, making measurements close to the fuel injector proved to be difficult. This was because the quantity of smoke in the region was greater than 1000mg/m3. It was found necessary to dilute the sample prior to the determination of the particle number concentration using SMPS. The issues associated with SMPS dilution are discussed.


1993 ◽  
Vol 115 (3) ◽  
pp. 554-562 ◽  
Author(s):  
L. H. Cowell ◽  
K. O. Smith

Development of a lean-premixed, liquid-fueled combustor is in progress to achieve ultra-low NOx emissions at typical gas turbine operating conditions. A filming fuel injector design was tested on a bench scale can combustor to evaluate critical design and operating parameters for low-emissions performance. Testing was completed using No. 2 diesel. Key design variables tested include premixing length, swirler angle, injector centerbody diameter, and reduced liner cooling. NOx emissions below 12 ppmv at 9 bar pressure were measured. Corresponding CO levels were 50 ppmv. An optimized injector design was fabricated for testing in a three injector sector of an annular combustor. Operating parameters and test results are discussed in the paper.


Author(s):  
K. D. Brundish ◽  
M. N. Miller ◽  
C. W. Wilson ◽  
M. Hilton ◽  
M. P. Johnson ◽  
...  

The objective of the work described in this paper was to identify a method of making measurements of the smoke particle size distribution within the sector of a gas turbine combustor, using a Scanning Mobility Particle Sizing (SMPS) analyser. As well as gaining a better understanding of the combustion process, the principal reasons for gathering these data was so that they could be used as validation for Computational Fluid Dynamic (CFD) and chemical kinetic models. Smoke mass and gaseous emission measurements were also made simultaneously. A “water cooled,” gas sampling probe was utilised to perform the measurements at realistic operating conditions within a generic gas turbine combustor sector. Such measurements had not been previously performed and consequently initial work was undertaken to gain confidence in the experimental configuration. During this investigation, a limited amount of data were acquired from three axial planes within the combustor. The total number of test points measured were 45. Plots of the data are presented in 2 dimensional contour format at specific axial locations in addition to axial plots to show trends from the primary zone to the exit of the combustor. Contour plots of smoke particle size show that regions of high smoke number concentration once formed in zones close to the fuel injector persist in a similar spatial location further downstream. Axial trends indicate that the average smoke particle size and number concentration diminishes as a function of distance from the fuel injector. From a technical perspective, the analytical techniques used proved to be robust. As expected, making measurements close to the fuel injector proved to be difficult. This was because the quantity of smoke in the region was greater than 1000 mg/m3. It was found necessary to dilute the sample prior to the determination of the particle number concentration using SMPS. The issues associated with SMPS dilution are discussed.


Author(s):  
Christopher M. Heath ◽  
Yolanda R. Hicks ◽  
Robert C. Anderson ◽  
Randy J. Locke

Performance of a multipoint, lean direct injection (MP-LDI) strategy for low emission aero-propulsion systems has been tested in a Jet-A fueled, lean flame tube combustion rig. Operating conditions for the series of tests included inlet air temperatures between 672 K and 828 K, pressures between 1034 kPa and 1379 kPa and total equivalence ratios between 0.41 and 0.45, resulting in equilibrium flame temperatures approaching 1800 K. Ranges of operation were selected to represent the spectrum of subsonic and supersonic flight conditions projected for the next-generation of commercial aircraft. This document reports laser-based measurements of in situ fuel velocities and fuel drop sizes for the NASA 9-point LDI hardware arranged in a 3 × 3 square grid configuration. Data obtained represent a region of the flame tube combustor with optical access that extends 38.1-mm downstream of the fuel injection site. All data were obtained within reacting flows, without particle seeding. Two diagnostic methods were employed to evaluate the resulting flow path. Three-component velocity fields have been captured using phase Doppler interferometry (PDI), and two-component velocity distributions using planar particle image velocimetry (PIV). Data from these techniques have also offered insight into fuel drop size and distribution, fuel injector spray angle and pattern, turbulence intensity, degree of vaporization and extent of reaction. This research serves to characterize operation of the baseline NASA 9-point LDI strategy for potential use in future gas-turbine combustor applications. An additional motive is the compilation of a comprehensive database to facilitate understanding of combustor fuel injector aerodynamics and fuel vaporization processes, which in turn may be used to validate computational fluid dynamics codes, such as the National Combustor Code (NCC), among others.


Author(s):  
Hu Li ◽  
Mohamed Altaher ◽  
Gordon E. Andrews

Biofuels offer reduced CO2 emissions for both industrial and aero gas turbines. Industrial applications are more practical due to low temperature waxing problems at altitude. Any use of biofuels in industrial gas turbines must also achieve low NOx and this paper investigates the use of biofuels in a low NOx radial swirler, as used in some industrial low NOx gas turbines. A waste cooking oil derived methyl ester biodiesel (WME) has been tested on a radial swirler industrial low NOx gas turbine combustor under atmospheric pressure and 600K. The pure WME and its blends with kerosene, B20 and B50 (WME:kerosene = 20:80 and 50:50 respectively), and pure kerosene were tested for gaseous emissions and lean extinction as a function of equivalence ratio. The co-firing with natural gas (NG) was tested for kerosene/biofuel blends B20 and B50. The central fuel injection was used for liquid fuels and wall injection was used for NG. The experiments were carried out at a reference Mach number of 0.017. The inlet air to the combustor was heated to 600K. The results show that B20 produced similar NOx at an equivalence ratio of ∼0.5 and a significant low NOx when the equivalence ratio was increased comparing with kerosene. B50 and B100 produced higher NOx compared to kerosene, which indicates deteriorated mixing due to the poor volatility of the biofuel component. The biodiesel lower hydrocarbon and CO emissions than kerosene in the lean combustion range. The lean extinction limit was lower for B50 and B100 than kerosene. It is demonstrated that B20 has the lowest overall emissions. The co-firing with NG using B20 and B50 significantly reduced NOx and CO emissions.


Author(s):  
Kenneth O. Smith ◽  
F. R. Kurzynske ◽  
Leonard C. Angello

The design and testing of three natural gas fuel injector configurations for a low emissions gas turbine combustor are described. The injectors provided varying degrees of fuel/air premixing and permitted an assessment of the degree of premixing necessary to achieve NOx emissions below the program goal of 10 ppm. The work described represents a preliminary step in an effort to develop production-level gas turbine combustor hardware with ultra-low NOx capabilities.


Sign in / Sign up

Export Citation Format

Share Document