OCR Text |
Show c Q) Ol a ~ .~ O.B Z 0.6 -ocad a c 0.4 a ·u ro '- LL o 65 ................. ...... -'v.. ~ :.... T "'T ~ .. .. "" ···· ... T T Yl. '- V " T "-"',- '~.~ --e--------~ . - - - - 6- -1:1. :l-.&.-H-&.C_N ............ .-.I '. Tar 70 75 80 85 90 95 Carbon Content. daf wt 0/0 Fig. 4. Comparison between predicted and observed char-N (solid curve and 0 vs .), tarN (dotted curve and V vs"), and HCN (dashed curve and 0 vs .) from several coals for devolatilization during heatup at more than 104 K/s vs the daf carbon content of the samples. probably faster than coal particles can actually be heated in wire mesh heaters. Note that the equal observed values for 1000 and 5000 Kls are consistent with this rationale. Both of the predictions correctly grow for faster heating rates, especially for tar-No Tar-N increases in proportion to the tar yields which increase for faster heating rates. Coals from across the rank spectrum are considered in the evaluation in Fig. 4, based on the ultimate nitrogen evolution data reported by Chen and Niksa [8]. Each case is for primary devolatilization during heatup at rates in excess of 104 K/s to the point where asymptotic weight loss was achieved, so there were no isothermal reaction periods. Predicted char-N levels correctly exhibit the minimum nitrogen retention observed for hvA bituminous coals, and are within experimental uncertainty for all coal types. Predicted HCN levels are insensitive to coal rank for the short contact times in these experiments, and within experimental uncertainty for all but the highest ranks. The overprediction for the subbituminous coal is easy to explain, . because the data show a sizable breach of closure in the nitrogen balance which is probably due to NH 3 production. The predicted levels of tar-N correctly. exhibit a maximum for hv bituminous coals, consistent with the maximum tar yields for coals of this rank. 6 |