Studies of the Release of Sodium from Pulverised Coal in a Flat Flame Burner

Abstract

This paper investigates the problem of the rate of release of alkalis from coal/coal blends as they are burnt in Utility Boilers with the aim of evolving better models of the effect of given fuel blends on slagging and fouling in real boiler situations. Although numerous slagging and fouling indices have been evolved over many years to describe the behaviour of coals in given boiler situations they often fail to give the required results and prove to be of little use when co-firing blends of coal with biomass materials. In this work we have constructed a natural gas flat flame burner with associated furnace and viewing ports so we can use a PLIF system to look at the evolution of sodium from burning coal/coal blend particles and relate this to the condition of the ash in the burning coal particle as the ash particle reaches the region of the superheater tubes in a boiler. By appropriate adjustment of the air fuel ratio we can produce temperature/residence time distributions downstream of the flat flame burner and in the furnace section which closely parallel those found in Utility Boilers. Pulverised coal particles as produced for boiler firing are sieved into two fractions, those in the size range 67 to 37 microns and above 67 microns (virtually no residue was found below 37 microns). This was to ensure that the finer fraction had a narrow size distribution, which devolatilised and burnt in a time frame commensurate with that of our system Particle velocities could be readily measured by streak velocity using the available PLIF camera via their radiation signal produced from their combustion process. The camera was aligned such that the initial devolatilisation process was not observed as the associated radiation could mask the PLIF image; this proved to be satisfactory as significant sodium evolution did not occur until after the end of the devolatilisation process. PLIF images of the sodium evolution were generated via excitation of the D₂ line at 16960.95 cm⁻¹ with collection via a gated intensified CCD camera (Princeton Instruments ICCD- 576). Four groups of images (500 images per set) were generated and collected at each setting to ensure accuracy of results, with a polarisation filter to minimize interference from Mie scattering with the sodium fluorescence. Novel results have thus been produced which will allow the evolution of better models for predicting slagging and fouling, especially when the work is extended to other volatile species in coal ash.