OCR Text |
Show 6 model of gas-particle transport processes [Hurt and Mitchell, 1992] was used to relate this temperature data to global char reactivities at various conversions [Hurt and Davis, 1994]. Reactivity is shown to decrease monotonically during combustion, dropping by a factor of 5 at 90% conversion. Global kinetic model assumptions, typically based on a constant, conversionindependent global reactivityt do not lead to large errors below 50% conversion, but, by 70% conversion, the reactivity has decreased by a factor of about 2 and numerous near-extinction events are observed (see Fig. 2). The data for the lignite and the biomass chars is, again, very different in terms of extinction phenomena. The optical data from these three types of particles does not display any evidence of rapid reactivity loss at intermediate extents of burnout. For example, even after Beulah lignite has reached a bulk carbon conversion of 90 percent, virtually all carbon-containing particles remain at the high temperatures characteristic of the low conversion regime. This experimental work on laboratory chars has been complemented using the same entrained flow reactor and optical diagnostic to evaluate the reactivities of residual carbon in boiler flyash. A suite of fly ash samples was obtained from commercial scale industrial and utility boilers and a technique was developed to separate and extract residual carbon from the inorganic portion of the fly ash [Hurt and Gibbins, 1994]. The results of these reactivity tests are summarized in Fig. 3. To facilitate comparisons, reactivities are presented as burning rates in a selected environment and are plotted vs. rank of the parent coal, represented by its dry, ash-free carbon content. For comparison to the residual carbon materials, Fig. 3 also plots the reactivities of 12 different laboratory-generated chars at low-to-intermediate conversion [Hurt and Mitchell, 1992]. The bars around the residual carbon data points represent the range of reactivities determined using different assumptions regarding the CO/C02 ratio at the particle surface [Hurt and Gibbins, 1994]. Char reactivity is seen in Fig. 3 to be a strong function of rank, with the reactivities of the laboratory chars falling in a well-defmed band as a function of parent coal carbon content. At any given rank the residual carbon samples are less reactive than the laboratory chars by factors of 2 to 7. Also, in each case where chars were available from the same parent coal, the residual carbon samples are seen to be much less reactive. It is also found that thennally pretreating the samples at 1000 °c for 15 minutes to remove surface oxides has little affect on reactivity. Despite the t An exception is the recent model of Charpenay and coworkers [Charpenay et al., 1992], which bases char reactivity on instantaneous char hydrogen content and thus has the potential to describe some conversion dependence. However, as will be discussed, this model has difficulties in predicting the reactivity loss that we observe. |