Abstract

One of the main objectives of combustion research in field of gas turbine application during the last decades was and still is the reduction of pollutant emissions. The most promising technology to reduce these pollutants turned out to be Lean Premixed (LP) and Lean Premixed Pre-Vaporized (LPP) combustion. However, serious problems concerning combustion-driven instabilities occurred with the implementation of the LP/LPP-concept. Today, prediction and systematic suppression of self-sustained combustion instabilities is an issue still unsolved, due to incomplete understanding of the physical feedback mechanism and the lack of models for dynamic flame response, i.e. frequency dependent characteristics of LP/LPP swirl flames. In that context, the purpose of the current paper is the establishment of a physical model to describe frequency dependent flame dynamics concerning burning velocity of steady-state premixed flames. Derived from that basic understanding, scaling laws for the prediction of unstable operation conditions will be established in dependence on main operation parameters such as thermal load, mixture temperature, air equivalence ratio and especially of fuel and operating pressure. Therefore, a new swirl-burner has been designed, offering the feasibility to choose the type of fuel, to adjust the swirl number for main and pilot burner and the burner exit geometry steplessly and to vary preheating temperatures, air equivalence ratios and thermal loads in a range of industrial relevance for gas turbine applications. To establish a periodical modulation of the mixture mass flow of the main L(P)P flame at the burner outlet sinusoidally in-time with well-defined frequencies and amplitudes, a pulsating unit was used. Using a mixing/ pre-vaporizing unit to create a time-independent and spatial homogeneous mixture of natural gas/ kerosene vapor and combustion air at the burner outlet, flame transfer functions of LP- and LPP swirl flames depending on main operating parameters were determined. The discussed results then lead to stability map for a given combustion system depending on the main operation parameters based on the knowledge of only one fully-described parameter combination leading to an instable condition. Based on this scaling procedure and confirmed by further experimental work the prediction of stability limits depending especially on the type of fuel, the swirl number and the operating pressure will be possible.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call