This paper describes the development phases of a damping system for combustion instability reduction in an annular type combustor for heavy-duty gas turbine applications. As reported by the authors in a previous paper, the full scale annular test rig allowed for an extensive characterization of the combustor with realistic acoustic boundaries at engine-relevant conditions. Emissions and operability assessment over a wide range of load conditions was performed, allowing the evaluation of the response of the system near the thermo-acoustic instability onset.

The instability is quantified by its acoustic growth rate. This quantity is a crucial input in the design process of dampers. A methodology has been used to extract these growth rates form measured pulsation data.

Experimentally determined growth rates have been evaluated for different fuel flow rate split between the main and the pilot injections, providing input to dampers preliminary design. Given current combustor architecture constraints, a first attempt configuration has been proposed and performances evaluated in the full annular rig. Dampers have been equipped with dynamic sensors and thermocouples with the purpose of measuring the growth rate abatement and the consequent NOx emissions reduction. A dedicated numerical toolbox, in-house developed by GE Power, has been used for both dampers preliminary design and growth rate reduction evaluation. Fine tuning of dampers elements as well as design assumptions adjustments required additional experimental evaluations and design iterations.

Encouraged by the successful test in the concept phase, an optimized design for engine implementation was defined, that featured a significant increased damper volume, involving combustor parts re-design. The optimized configuration was finally tested in full annular rig and results demonstrated an important enhancement of operability while maintaining NOx emissions below the target levels.

This content is only available via PDF.
You do not currently have access to this content.