OCR Text |
Show boundary conditions, with vorticity surfaces shedding from all corners in the furnace interior. The turbulence level in the inflow stream entering the secondary furnace is simulated by vortex sheets. As in all LIM simulations, the initial conditions correspond to the onset of inflows into the initially quiescent furnace. 4.2 Predicted natural gas rebum performance LIM simulations were conducted for natural gas injection at the 208-ft. and 229-ft. elevations for each of the injector combinations in the test matrix. Examples of typical gas mixture fractions fields that result are shown in Figs. 16 and 17. These show the instantaneous vorticity surfaces, as well as the mixture fraction fields £(x,f) and molecular mixing rate fields V^V^(x,/). Owing to the relative homogeneity of the oxygen, temperature, and NOx fields within the Joliet Unit 6 furnace, the LIM simulations results show that there is virtually no effect of either the injector elevation or the choice of injector groups on the level of NOx reduction achieved at any given level of gas input to the boiler. The simulations further show that the % NOx reduction increases virtually linearly with the % gas heat input to the boiler. with no limit in the NOx reduction being reached over the range of gas input values simulated. The LIM-based simulations predicted 36-38% NOx reduction at 7 % gas heat input to the boiler, with the slight variation reflecting the effects of the different gas injector groupings and elevations. The simulations also predicted CO emissions of 188-214 ppm at this level of NOx reduction. 4.3 Comparisons with measured reburn performance Field testing of the fuel-lean gas reburn system performance at Joliet Unit 6 was begun after all the LIM simulation results were reported. Data for NOx and CO emissions were obtained from the tests at various gas heat input levels to the boiler for various gas injector groupings and elevations. Test results obtained to date were recently reported by Glickert etal (1997). and cover the reburn system performance at full load (320 MWe) conditions. These results are shown by symbols in Fig. 18. with the results predicted by the LIM-based simulations shown by lines in the same figure. CO comparisons are shown in Fig. 19. Note that the LIM simulations correctly predicted all results for the fuel-lean gas reburn performance available to date from Joliet Unit 6 under no-steam injection conditions. The simulations predicted slightly more than 5 % NOx reduction for each 1% gas heat input to the boiler, and this is verified by the field data. The field data also confirmed the prediction that the gas injector groupings and elevations would have no measurable effect on the level of NOx reduction achieved. The data further confirmed the prediction that 36-38% NOx reduction would be achieved at 7 % gas heat input. Finally, as regards C O emissions, while the field tests showed considerable scatter in the measured CO level, the 188-214 ppm emissions predicted by the LIM simulations appear to be in good agreement with the measured performance. 6. CONCLUDING REMARKS LIM-based simulations were performed for fuel-lean gas reburning in two very different coal-burning utility boilers to obtain an assessment of the accuracy of such |