OCR Text |
Show 7.2.3. The Series 3-XBM2-F three-burner, full-furnace trials In the single-burner trials described above, the floor sink panels were fully covered with 25 m m thick blanket insulation, and the heat flux through these was small, resulting in high temperatures in the combustion chamber. In the full-furnace three-burner trials, the floor coverage was widely varied, Fig. 4, and the furnace temperature ranged from very high down to below the minimum at which stable combustion could be maintained. The results on N O x emissions in the five trials with stable combustion agree with (5), but the ranges of C0 and n are greater than in the single burner trials, because of the greatly enlarged furnace temperature range. Figure 7 shows the graph of the exhaust-gas N O x level at X0 = 3 % and Ta = 400 °C vs. the firing rate. The coefficient C0 is the value of the ordinate in Fig. 7 at mf(-Ahc) = 900 k W (3 burners at 300 k W each), divided by 1.80. The fitted values of C0 and n are given in Table 5. 7.2.4. The Series 1-XBM2-F single-burner, full-furnace trials In these trials, three levels of coverage of the floor sink were used, providing a range of furnace temperature levels, but the firing rate was set at the maximum and the exhaust-gas oxygen level, X0 , was maintained, as nearly as possible, at 3 %. The main point of these trials was the mapping of the gas composition field in order to determine the flame size. However, useful data were also obtained on N O x emissions. These are considered in the next section, § 7.2.5. 7.2.5. The effects of furnace-gas temperature level The fuel and air jets from the burner, and the enlarged jet that coalesces out of these when they meet, all entrain recirculating furnace gases (drawn into recirculation by that very entrainment). These recirculating combustion products, which have lost temperature through furnace heat transfer, must considerably affect the temperatures reached in the reaction zone and thus the emissions of N O x . The temperature level of the recirculating furnace gases is thus an important factor in N O x production. The most readily available measure of this quantity is the exhaust gas temperature, Te. Figure 8 shows as a function of Te the exhaust-gas N O x level at X0 = 3 %, Ta = 400 °C and m/(-A/7c)=300kW per burner for the Series 1-XBM2-T, 3-XBM2-F and 1-XBM2-F trials. The results for the single burner trials, comprising those with the tunnel combustion chamber and the full furnace chamber, are plausibly fitted by a single curve. The results for the three-burner trials show a similar temperature dependency but nearly a doubling in level. The common temperature dependency approximates XUOi oc Te 46, where Te is in kelvins. Some additional information, of interest in characterizing the operating regime, is included in the figure. Curve a represents the calculated N O x levels in the combustion products at chemical equilibrium at temperature Te. Curves b and c represent the N O x levels predicted treating the combustion chamber as a perfectly stirred reactor operating at Te, using the simplified Zeldovich mechanism and assuming that atomic oxygen, O, is in equilibrium with molecular oxygen, O2. Clearly the observed N O x emissions are far below the lowest possible equilibrium levels, curve a. O n the other hand, at Te < 1600 K they are far above the predictions of the well-stirred reactor model for 10 |