OCR Text |
Show However, under such conditions the reburn gas penetrates into the high oxygen-containing flow in the back half of the furnace, where it is rapidly oxidized and makes little contribution to NOx removal. A more preferable scenario would be for the gas jets to penetrate only up to the oxygen mixing layer, where oxygen concentrations are relatively low and NOx concentrations are reasonably high, and thus significant reburning can occur before the C H radicals are depleted. The simulation results show that it is for this reason that the downward injection nozzles gave better reburn performance than did the straight injectors, since they place the gas downward into the low oxygen containing front region of the furnace and thus avoid penetrating through the oxygen mixing layer. Similarly, it is due to this oxygen mixing layer that injecting more gas - and thus achieving higher penetration - did not lead to higher NOx removal effectiveness. Lastly, this is also the reason why introducing gas through the sidewall injectors produced less effective reburn performance than did the front wall injectors alone, since the two rearmost pairs of sidewall injectors place gas directly into the high oxygen-containing back portion of the furnace. The effect of gas injection at different furnace elevations but at otherwise comparable conditions is also shown in Figs. 10 and 11. These give typical instantaneous injected gas mixture fraction and molecular mixing rate fields that result from injection at the 6th, 5th. and 4th floor elevations in the furnace. Owing to the overall gas path that results from the roof-fired configuration, it can be seen that injection at the 6th floor elevation gives the most advantageous distribution of gas within the furnace. Examples of the accompanying reburn chemistry are shown in Figs. 12 and 13. where individual chemical species histories of Lagrangian particles are shown, as well as the resulting reburn efficiency index M. The combined effect of these mixing and chemistry considerations are reflected in the measured reburn performance shown in Fig. 14. The data obtained at 0 % SOFA settings in Fig. 14a show that the % NOx reduction increases roughly linearly with the % gas heat input to the furnace. In the absence of SOFA air. there is no mixing layer that separates a fuel-rich zone near the front wall from an oxygen-rich zone in the back part of the furnace. As a consequence, there is no limit found in the NOx reduction achieved as the gas input to the furnace is increased over the range of values tested. On the other hand for the relatively high SOFA settings in Fig. 14b. the % NOx reduction is initially linear and has a higher slope at low gas heat inputs, owing to the lower oxygen concentrations that result near the front wall at high SOFA settings. However with increasing % gas heat input to the furnace, the gas jets eventually penetrate through the mixing layer and further gas input leads to little or no further NOx reduction. The limiting % NOx reduction is seen to increase with higher SOFA setting, as would be expected from the underlying physical mechanism that was revealed by these numerical simulations. 3.5 Recommendations for improved reburn performance Based on the insights obtained from the LIM simulations of the existing fuel-lean gas reburn system at Elrama Unit 2, the following recommendations were made to achieve increased levels of NOx reduction at lower levels of gas heat input. All of these recommendations are based on exploiting the fuel-rich region near the furnace front wall, and avoiding the high oxygen region found in the back half of the furnace. |