OCR Text |
Show 60 cf!. 50 L- a 30 20 • • ""\ / \ / \. /e--•~ 13" • / / \ ~Ii\\. ~• /. ,)IJ'b / 8"\ rI 'lY •• 'b..,1 • ~ • 78 80 82 84 Carbon Conlent, wt % daf • • !J6 Fig. 1. Reported ultimate yields of volatiles (upper e) and tar yields (lower.) for atmospheric pyrolysis ofhigh volatile bituminous coals with heating rates greater than 1000 KJs. FLASHCHAIN predictions for total and tar yields appear as the open squares connected by solid and dashed lines, respectively. The series of 4 data evaluations in Fig. 2 cover a wider range of coal rank and also demonstrate the predicted devolatilization histories from onset to completion. These cases are for isothermal reaction periods of 4 s following heatup at 3000 KJs to the indicated temperatures. In so far as ultimate yields are not achieved unless temperatures exceed 1100 to 1200 K, depending on coal type, these data express the reaction dynamics in terms of a temperature dependence. The series is arranged in order of increasing rank of the samples to illustrate several facets of the rank dependence as continuous variations. FLASHCHAIN correctly predicts that (1) The onset of devolatilization shifts to higher temperatures for coals of higher rank; (2) The initial weight loss from low rank coals is dominat~d by gases, whereas tars are the predominant early product from bituminous coals; (3) Tar evolution ceases before gas evolution; (4) tntimate weight loss is fairly constant for ranks through hv bituminous, then falls off markedly for coals of higher rank; (5) Tar yields pass through a weak maximum for hvA bituminous coals; and (6) The fraction of the total weight loss that is tar is an increasing function of rank. Specifically, the FLASHCHAIN predictions for weight loss are within experimental uncertainty in all 3 . |