| OCR Text |
Show COMPARISON TO INTERMEDIATE·SCALE AND FULL·SCALE A task in the 10w-NOx cell burner project and a preliminary study for the cyclone reburning project investigated field constraints for retrofits. These constraints included structural steel placement and windbox design, among others. The studies set conditions for available space to run air or coal lines, locate reburners, etc. In tum, this infonnation was used in the design of the pilot-scale test programs. Cell Burner Project Description Six-million Btulhr pilot-scale and IOO-million Btulhr intennediate-scale tests were conducted in the CFPF and L WS facilities, respectively. As noted, the results of those studies have been reported elsewhere. Detailed comparison of the results in tenns of the scale up to intennediate and full-scale will be addressed here. Figure 4 shows schematics of the two test furnaces drawn to scale. Figure 6 shows side views of the pilot-, intennediate-, and full-scale units for comparison. While the CFPF and LWS are of similar design, some subtle differences explain minor variations in the test data. The LWS, even with refractory in the burner zone, probably has higher heat transfer to its walls (than the CFPF), resulting in lower flame temperature and FEGT. Firing at 100- million rather than 150-million Btulhr would also tend to lower both. Firing at IOO-million Btulhr resu]ts in a lower furnace liberation rate (8.7 versus 17.1 Btulhr/ft3) than for the CFPF firing at 6-million Btulhr. The CFPF rate is closer to that for the DP&L unit (19.9 Btu/hr/ ft3). NOx is a function of flame temperature. Burner zone heat transfer will affect this temperature. Residence time is closer to full-scale for the CFPF, too (3 versus 4 seconds). Net heat absorbed per adjusted cooled surface area (HNSC) is an accepted parameter used for predicting NOx in utility boilers. Trying to adapt this parameter to boiler simulators like the CFPF and LWS is not useful. The L WS burners were scaled up versions of the CFPF ones, but a 2x2 array was used to more closely simulate field conditions. Thus, the CFPF project used two 3-million Btulhr burners and the L WS used four 25-million Btulhr burners. Both projects used single wall firing. The two test coals were deliberately chosen to have similar nitrogen content, but widely different volatile matter. Thus, parameters such as fixed carbon to volatile matter of importance in burner emissions perfonnance were also different. Tables 3 and 4 show the respective coal analyses for the two test programs. It can be seen that there was some variation between the coals purchased for the two test programs for the same nominal parent coal. This happened in part because the coals were bought at different times. Fuel grind was nominally the same for both pilot programs at 70% through 200 mesh. The average coal analysis for the DP&L demonstration test is shown in Table 5 for comparison. This represents the average properties of the fairly tight range of coals which will be used during the demonstration program. 6 |
