Experimental Study on Pressurized Oxyfuel Combustion of Coal s2

Flame Development in Pulverised Biomass for Fine Particles David J.F. Slatter, Muhammad Azam Saeed, Clara Huéscar Medina, Gordon E. Andrews*, Herodotos N. Phylaktou and Bernard M. Gibbs Energy Research Institute, School of Chemical and Process Engineering, University of Leeds, Leeds, LS2 9JT, UK

Abstract for oral presentation: Pulverised biomass is a significant source of carbon neutral renewable electricity and hence plays an important role in carbon reduction strategies. This biomass is burned either co-fired with coal or increasingly as 100% pulverised biomass. The biomass particle size fed to burners is relatively coarse with little of the powder <63µm and most in the 63 to 1000µm size range. There is little information on how these relatively coarse biomass particles propagate flames in boilers or in their associated dust explosion hazards and this work presents high speed videos of flame propagation in coarse particle biomass dust clouds as a means of deducing the mechanism of flame propagation. There is little in the open literature on flame propagation and explosion risk for fine biomass particles and this work presents some results in this area using the Hartmann dust explosion tube equipment with high speed photography to follow the flame development. The equipment was fitted with a vertical line of three thermocouples (red lines) to determine the flame speed, which was also determined from high speed video recordings. Lycopodium, and oak sawdust were investigated; lycopodium powder is a naturally occurring biomass that is ultra-fine with no requirement for additional milling. The oak sawdust was sieved to ≤63µm. The flame propagation measurements in the Hartmann explosion tube showed that the uniformity of the flame front correlated with the particle size of the fuel’s tested. For fine lycopodium powder the flame propagated smoothly and behaved in a similar way to a gas flame. Similar results were obtained for oak milled to <63µm. The flame speed for the two materials were 3.45m/s for lycopodium and 2.99 m/s for oak less than 63 microns. The flame speed was shown to be controlled by the particle size and the morphology and chemical differences in the two particles had little influence on the flame speed. For rich mixtures the initial flame front before the Hartmann vessel vent burst is the only flame that propagates at the injected concentration. Once this flame bursts the vent, the flame exits the top of the tube and then there is a second flame in the tube that is burning dust not burnt in the first flame due to the lack of air. This flame was faster than the first flame. Air then continues to diffuse in and mix with the unburnt dust until there was a third flame and this has a lower flame speed due to leaner mixtures.

Keywords: dust explosions, flame propagation, particle size, flammability limits

Acknowledgement: The authors are grateful to the EPSRC Energy Programme (Grant EP/H048839/1) for financial support of part of this work. David Slatter would like to thank his parents for funding his PhD research.

* Corresponding author: e-mail: [email protected] Tel: 441133432493