OCR Text |
Show CL [Linby, 10J K/s. 2 s hold. 1 atm C ' ' i " i c 0) y 0. £ 'Gedling, 103 K/s, 2s holdTTaTm 120 k 100 « 60 L c O 40 OJ 03 20'. I Pit 8, 103K/s, 2s hold, 1 atnT 0 I i 1 1 L_ , 1 ___- 400 600 800 1000 1200 1400 1600 Temperature, C Figure 1. Evaluation of predicted char compositions against datasets of Cai (1995) for 3 hv bituminous coals, including fractions of the coals' original levels of C(B), H(«), 0(dotted line), and N(A), plus the fractional char yields(O). Secondary Volatiles Pyrolysis C F D simulations of full-scale, coal-fired utility furnaces are now being used for troubleshooting and design worldwide However, one recent literature survey (Niksa 1996) noted universally poor predictions of gaseous emissions m the exhaust, with no instances of quantitative agreement within useful tolerances for C O and N O concentrations anywhere in furnaces. This important limitation was attributed to deficiencies in the submodel for combustion of volatile matter. Almost universally, the mechanism that is utilized in current C F D simulations asserts combustion of volatiles at their rate of mixing with air according to eddy breakup or eddy dissipation mechanisms. Equilibrium combustion products are usually assigned based on the assumption that volatiles have the same elemental compositions as their parent coals. The mixing-limited approach to equilibrium combustion products is implausible for two reasons: (1) Volatiles have markedly different compositions than their parent coals because dunng rapid coal devolatilization. all the coal's oxygen and sulfur and 80 to 9 0 % of its hydrogen are released, whereas only about half trie carbon and nitrogen are expelled into the gas phase. (2) Products of primary devolatilization are rapidly transformed by secondary volatiles pyrolysis once they have been expelled into not gases. Consequently, the very complex distributions of chain and condensed polynuclear aromatic hydrocarbons, carbon oxides, water and hydrogen that are released initially - and would appear to burn very rapidly - are converted into much simpler mixtures of soot, carbon oxides. H 20. and H2 with small amounts of C H 4 and C 2 H2 (Bruinsma, 1988, Nelson 1986; Chen, 1992). The abundance of soot. CO, and stable combustion products in the secondary pyrolysis products ensures that they will burn relatively slowly (Mariow 1992; Cho, 1995). Indeed, soot and C O may not even compete effectively for the available oxygen with the especially reactive chars from coal ranks of subbituminous and below (Niksa, 1998). The primary product distributions from FLASHCHAIN'" are converted to secondary pyrolysis products by applying phenomenological mechanisms based on the observations of Chen et al. (1992). The distributions are formulated in terms of an extent of secondary pyrolysis, so no finite-rate kinetics are incorporated. The extent of secondary pyrolysis is assigned as the soot fraction, which is the ratio of the yield of soot to the sum of the yields of tar, oils, and soot The soot composition is based on a constant value of H/C for all soots from all coal types, and retention of a fixed fraction of the nitrogen in tar All heteroatoms m tar are expelled as noncondensibles. Tar-0 is released as C O n exchange for incorporation of C 2H2 into soot. Tar-N s released as H C N and tar-S is released as H2S The bulk of tar-H is released as H2. Consequently, the sum of the yields of tar, soot, and oils remains nearly mvanant throughout secondary pyrolysis, even though the yields of CO. C2n2. H2 and H C N surge throughout. Non-condensible hydrocarbons are converted into C H 4 and C2H2, although C H 4 IS aisc converted into C2H2 during the last half of secondary pyrolysis. Product distributions based on this phenomenoiog:cai mechanism are evaluated against data in Fig. 2 The measurements were reported by Chen et al. (1992) for a P t 8 hv bituminous coal. The predictions in Fig. 2 aooear versus the soot fraction. Values range from zero, which denotes primary devolatilization, to unity, which denotes the complete conversion of tar into soot. In the laboratory progressively higher soot fractions were achieved by raising the furnace temperature. The faster heating -ates associated with progressively higher furnace temperatures were accounted for in the simulations. Predicted.weic.ht ess |