0
Research Papers: Gas Turbines: Combustion, Fuels, and Emissions

# Instability Control by Premixed Pilot Flames

[+] Author and Article Information
Peter Albrecht1

Hermann-Foettinger Institute (ISTA), Technical University of Berlin, D-10623 Berlin, Germanyalbrecht@pi.tu-berlin.de

Hermann-Foettinger Institute (ISTA), Technical University of Berlin, D-10623 Berlin, Germanystefanie.bade@pi.tu-berlin.de

Arnaud Lacarelle

Hermann-Foettinger Institute (ISTA), Technical University of Berlin, D-10623 Berlin, Germanyarnaud.lacarelle@pi.tu-berlin.de

Christian Oliver Paschereit

Hermann-Foettinger Institute (ISTA), Technical University of Berlin, D-10623 Berlin, Germanyoliver.paschereit@tu-berlin.de

Ephraim Gutmark

Department of Aerospace Engineering and Engineering Mechanics, University of Cincinnati, Cincinnati, OH 45221-0070ephraim.gutmark@uc.edu

1

Corresponding author.

J. Eng. Gas Turbines Power 132(4), 041501 (Jan 12, 2010) (8 pages) doi:10.1115/1.3019293 History: Received February 25, 2008; Revised August 25, 2008; Published January 12, 2010; Online January 12, 2010

## Abstract

Premixed flames of swirl-stabilized combustors (displaced half-cone) are susceptible to thermo-acoustic instabilities, which should be avoided under all operating conditions in order to guarantee a long service life for both stationary and aircraft gas turbines. The source of this unstable flame behavior can be found in a transition of the premix flame structure between two stationary conditions that can be easily excited by fuel fluctuations, coherent structures within the flow, and other mechanisms. Pilot flames can alleviate this issue either by improving the dynamic stability directly or by sustaining the main combustion process at operating points where instabilities are unlikely. In the present study, the impact of two different premixed pilot injections on the combustion stability is investigated. One of the pilot injector (pilot flame injector) was located upstream of the recirculation zone at the apex of the burner. The second one was a pilot ring placed at the burner outlet on the dump plane. A noticeable feature of the pilot injector was that an ignition device allowed for creating pilot premixed flames. The present investigation showed that these premixed pilot flames were able to suppress instabilities over a wider fuel/air ratio range than the conventional premixed pilot injection alone. Furthermore, it was possible to prevent instabilities and maintain the flame burning near the lean blowout when a percentage of the fuel was premixed with air and injected through the pilot ring. $NOx$ emissions were significantly reduced.

<>

## Figures

Figure 1

Test rig facility and the swirl-stabilized combustor: (1) laminar flow element, (2) pneumatic slide valve, (3) heat exchanger, (4+7) mass meter, (5+8+10) needle valve, (6) high voltage ignition system, (9) rotameter, (11) thermocouple for bottom exhaust temperature, (12) thermocouple for exit exhaust temperature

Figure 2

PFI installation at the cone vertex and Pilot Ring at the combustor's dump plane. Insert shows the prototype of the PFI: (1) sparkplug, (2) ground electrode, and (3) ceramic tube.

Figure 3

Stability map of a swirl-stabilized combustion for different preheat air temperatures of 300 K and 550 K and combustion outlet diameters of 65 mm and 200 mm. The ignition device was in all cases off, so that no PFI flame was generated.

Figure 4

OH-chemiluminescence and pressure signals for the case PFI-550 K at Φtotal=0.55

Figure 5

Power spectrum for the case PFI-550 K with and without PFI flame at Φtotal=0.55

Figure 6

Phase-averaged OH-chemiluminescence images at different phase shift angles at 200 kg/h main air mass flow rate and 300 K nonpreheated air for Φtotal=0.66. The instability occurred at 82 Hz.

Figure 7

OH-chemiluminescence images and rms pressure level at different operating points for the case PFI-550 K; the sparkplug was not activated. While for the curve “rich-lean,” the pressure fluctuations were investigated from a rich mixture to a lean mixture, and the rms pressure level was recorded for the curve “lean-rich” at the starting operating point Φtotal near lean blowout. A hysteresis effect was not clearly visible.

Figure 8

Time signal of the pressure fluctuations for the case PFI-550 K at Φtotal=0.55 and 0.61 with and without PFI flame. The frequency of the PFI flame was 110 Hz.

Figure 9

Time signal of the pressure fluctuations for the case PFI-300 K with and without PFI flame at Φtotal=0.66. The frequency of the PFI flame was 110 Hz.

Figure 10

Comparision of the OH-chemiluminescence between the side and central recirculation zones for the case PFI-550 K with and without PFI flame

Figure 11

rms pressure level for the case PFI-300 K with and without PFI flame. All rms values are normalized on the background noise of the baseline case at Φtotal=0.45.

Figure 12

NOx emission for the case baseline-550 K and for the case PFI-550 K with and without PFI flame

Figure 13

Influence of the pilot ring flames (diffusion ring and premixed ring) on the normalized rms pressure level. All rms values were normalized on the background noise of the baseline case at Φtotal=0.45.

Figure 14

NOx emission for different pilot ring injection cases

Figure 15

Mean flame heat release of the four injection configurations depending on the total equivalence ratio Φtotal

Figure 16

Impact of different pilot ring injection cases on the main flame stabilization

Figure 17

Mean flame heat release of the four injection configurations

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections