Fourth Australian Conference on Laser Diagnostics in Fluid Mechanics and Combustion The University of Adelaide, South Australia, Australia 7-9 December 2005 Study on atomic sodium release from pulverised coal particles in a Pre-mixed natural gas flame P.J. van Eyk 1,3 , C.Y. Wong 2 , N. Syred 4 , E.P. Ung 2 , Z.T. Alwahabi 1 , P.J. Ashman 1,3 and G.J. Nathan 2 Schools of Chemical 1 and Mechanical 2 Engineering, University of Adelaide, SA, Australia 5005 3 Cooperative Research Centre for Clean Power from Lignite 4 School of Engineering, Cardiff University, Wales, UK ABSTRACT Atomic sodium release during coal combustion has been shown to be a significant factor in fouling and corrosion of heat transfer surfaces within industrial coal-fired boilers. Although there has been significant research into the final forms of sodium from combustion, information about the release of sodium is limited. An initial study of the release of atomic sodium from pulverised coal particles in a pre-mixed natural gas flame was conducted using planar laser induced fluorescence, PLIF. Measurements were made of the intensity of sodium fluorescence, coal particle velocities and gas phase temperature within the flame in a purpose-designed laminar flame facility that provides similar time-temperatures history to coal particles in their early stages of combustion in an industrial coal-fired boiler. The velocity of the coal particles was measured to be about 6.1 m/s using particle streak velocimetry obtained from the natural luminescence of the hot particles and a 2.25 ms gate of the CCD camera. Temperatures were measured using thermocouples, and the maximum measured temperature was ~1500 °C. Calibration of the sodium fluorescence was not undertaken, but useful information was obtained of the release rate of sodium atoms from combusting pulverised coal particles. 1. INTRODUCTION The release of volatile alkali species during coal combustion is a significant factor in the fouling and corrosion of heat transfer surfaces within industrial coal-fired boilers. In particular, sodium and potassium species during coal combustion have been shown to form the initial deposit which accommodates subsequent build up of other ash particles. Sodium and potassium also form complex sulphate compounds that can lead to tube corrosion (Neville and Sarofim, 1985). Sodium chloride can also be deposited on boiler tubes when very high chlorine content coals are utilised (Steinberg and Schofield, 1996), leading to a higher potential for corrosion in these furnaces. The vaporisation and condensation of alkali compounds during coal combustion is not well understood. Whereas refractory oxides are vaporised and condensed in the early stages of the flame, volatile alkali constituents condense only in the cooler zones of the furnace, thereby allowing a sticky deposit to form on the boiler tubes (Neville and Sarofim, 1985). Various studies have been conducted to determine the principle initiators of heat transfer surface deposits in coal- fired boilers. The size distribution of the ash particles formed during pulverised coal combustion has been determined to be bimodal (Quann et al., 1982). Submicron aerosols of less than 0.1 µm are typically rich in alkali compounds, while larger (>1 µm) particles typically only have a small concentration of alkali. Hence there have been many studies on the alkali metal portioning in the ash from pulverised coal combustion (Neville and Sarofim, 1985, Gallagher et al, 1990). These studies have concluded that the fraction of the sodium condensing in the sub-micrometre fume is dependent on the amount of silicon in the coal and the temperature of the coal flame. The effect of silicon in the coal can be understood by the fact that the silicon compounds “capture” the sodium, thus not allowing the sodium to vaporise due to the low activity of sodium in the silica rich melt (Neville and Sarofim, 1985). Various of studies have been conducted to determine the effect of sulphur and chlorine on the deposit formation (Steinberg and Schofield, 1996), and the equilibrium conditions of the coal flame (Helble et al., 1992, Srinivasachar et al., 1990). Work undertaken by the State Electricity Commission of Victoria utilised atomic absorption to establish that the release of atomic sodium starts at 800 °C, and continues until above 1700 °C. The peak temperature of atomic sodium release was found to be about 1500 °C in both Loy Yang brown coal and North Dakota Beulah Lignite (Srinivasachar et al., 1990). Similar measurements using a Sodium Chloride standard revealed there is a large peak in the 800-900° C range. The absence of a similar low temperature peak for the coal samples suggests that the sodium in the coal was released as a form other than NaCl. While the final forms of sodium and the equilibrium conditions within the flame have been well investigated, there has been very little work done on the formation of the initial volatilised sodium atoms formed in the highly reducing environment within a flame. Particularly, the effect of concentration of the sodium species within the coal on the time scales of the volatilisation of the sodium atoms needs to be determined. One previous study undertaken at the University of Adelaide measured sodium emissions from coal particles burning in a natural gas Bunsen burner flame using planar laser induced fluorescence (PLIF) of sodium atoms (Nathan et al., 2003). The coal particles were supported by a ceramic alumina rod and held at the apex of the blue cone in the flame. The study showed that the sodium tends to be released at the time-scale of char combustion, rather than of the volatiles. The main limitation of the study for the present purposes is that the experiments used coal particles of approximately 3.5 mm, and hence a direct comparison with industrial boilers is not possible. Hence the present investigation aims to extend this 133