OCR Text |
Show quench rate provided a better NO control but with higher reconversion. However, in other experirrents with different delay times between rrethanol injection and entering the high quench zone, it was possible to achieve higher NO control perfonnance at higher quench rates. Obviously, NO-toN02 conversion can be optimized for specific application with a certain temperature profile. At optimum injection temperatures between 60 and 76% NO removal depending on temperature profiles can be achieved by injection of H20/CH30H reagents into coal flue gas at reagent/NO = 1.5. To evaluate low temperature N02 reconversion, tests were conducted with natural gas and coal firing in which sampling was performed at temperatures of 594, 477, 413, and 355 K that correspond to residence times of 0.8, 2.8, 6.2, 9.8 s, respectively. With both natural gas and coal, reconversion was not measurable below 594 K. These results show that if it is possible for N02 to survive to about 600 K, it will not be affected by reconversion and will reach the scrubber. KINETIC MODELING AND DISCUSSION The process of H20/CH30H injection for NO removal was modelled by using the CHEMKIN-II kinetic program developed by Kee et ale (1989). The mechanism GRI-Mech 1.1 (1994) was selected as a basis for describing H2, CO, CH20, CH4, ~H4' ~H6' and CH30H chemistry. The mechanism developed by Glarborg et ale (1993) was used for predicting chemical behavior of nitrogen oxides at high temperatures. However, the Glarborg's mechanism was designed to model the reburning process, and the scheme includes several species which do not participate in the H20 2-CH30H reactions. Reactions of the following 8 species were removed from the Glarborg's mechanism: HCN, HNCO, CN, ~N2' HOCN, NCN, NCO, and HCNO. On the other hand, several species and their reactions were added. A total of 6 species including S02' S03' HS03, HN03, CH30 2, and CH30 2H involving 20 reactions were added to the combined GRI-Mech/Glarborg 10 |