Laser-Induced Fluorescence Imaging in a 100 kW Natural Gas Flame

A tunable excimer laser at 248 nm (KrF) and 193nm (ArF) has been used to monitor two-dimensional OH and NO distributions in the turbulent flame of a 100 kW natural gas burner. Spatially resolved fluorescence (spatial resolution better than 1.0 mm) from a 20 cm x 20 cm area is collected under single shot conditions. We describe the problems encountered when laser-induced fluorescence imaging techniques are applied to large scale flames. Special experimental arrangements, imposed by the turbulent behavior of the flame we used, are also described. PACS: 07.65, 82.40.Py, 42.80 During the last three decades the reduction of pollutants in combustion processes has received great interest [1]. Different diagnostic methods, e.g. emission spectroscopy [2] and the line-reversal method [3], have been applied in order to characterize temperatures of flames mad densities of combustion species. About 15 years ago laser diagnostics entered the field of combustion with techniques as laserinduced fluorescence (LIF) [4] and coherent anti-stokes raman scattering (CARS) [4]. Recently other non-linear techniques such as degenerate four-wave mixing [5, 6] have been added to this list. Rayleigh scattering [7] and spontaneous Raman scattering in combination with LIF [8] are promising spectroscopic tools for flame diagnostics. So far, most techniques have been well experimented on at laboratory scale flames. Less experience has been obtained with the application of these techniques to larger scale flames, as for instance to flames in utility boilers of power plants. In the Netherlands all power plants together are responsible for 15% of the total NOx release in the Dutch atmosphere [9]. The combustion process in a utility boiler of a power plant, and its ahnost inevitable exhaust of nitric and sulfuric oxides, is mainly determined by the characteristics of the flames, such as temperature, species concentration distributions and local stoichiometry. The visualization of nitric oxide formation in the combustion process is of utmost importance for determining the optimum condition of the burner parameters, such as swirl and stoichiometry in order to develop a better, that is to say a cleaner, low NOx-burner [10]. Laser-induced (predissociative) fluorescence (LI(P)F) is one of the most sensitive diagnostic techniques used in combustion, as has been demonstrated by various groups [1116]. The detection of laser-induced predissociative fluorescence of OH [17] has an important advantage over detection of "normal" LIF of OH, because of its lack of collisional quenching at atmospheric pressures. When saturation of the electronic transition is avoided [18], LIPF gives the possibility of measuring densities and temperatures quantitatively, even in large two-dimensional areas. When OH is excited to the A2Z+(v I -3) state with a tunable KrF excimer laser at 248 nm, the off-resonant fluorescence to unoccupied states is monitored, so there are no problems of fluorescence reabsorption [17]. The excimer lasers can work at a repetition rate of 50 Hz and the imaging of fluorescence with highly sensitive UV cameras with gated image intensifiers yield the opportunity to perform on-line measurements, e.g., changes of the burner configuration or of the burner conditions on the NO distribution are directly observed on the video monitor. In this article we present two-dimensional laser-induced fluorescence measurements in a 100kW natural gas flame. We have succesfully monitored the OH and NO distributions in a large area with a tunable excimer laser operating on KrF and ArF, respectively, in a single laser pulse. This article describes, from an application-oriented point of view, the problems encountered when laser-induced fluorescence imaging techniques are applied to large scale flames. The complicated experimental conditions, among others caused by the turbulent behavior of the flame we used, forced us to improvise and to improve our experimental apparatus. These special arrangements are also described. 1 Experimental Setup The experimental setup is schematically given in Fig. 1. The different parts will be discussed below. Laser-Induced Fluorescence Imaging 165