OCR Text |
Show 1500 K gas stream containing 1% 02 at high and low rates of devolatilization. Figure 2 shows that the particle temperature history of a particle in an inert environment differs by less than 3 % from the particle temperature history in 1% 02. The difference in volatile release is shown in Figure 3; simulations of the particle in an inert environment shows insignificant differences from simulations of a particle in a 1% 02 environment. COAL COMBUSTOR SIMULATIONS The two-dimensional pulverized coal gasification and combustion code (PCGC-2) developed at Brigham Young University by Smith et al. (1980) was used to simulate a swirling coal combustor being studied by Hughes (1985) at the Canada Center for Mineral and Energy Technology ( C A N M E T ) and a non-swirling coal combustor at B Y U (Thurgood, 1979). This simulator is not presently capable of using integro-differential devolatilization rate equations, but the two sets of kinetic coefficients for the 2-step devolatilization model were used in the simulations. Details of the equations used in PCGC-2 are given by Fletcher (1983). The conservation equations of mass, momentum, and energy for both the gas and particle phases are modified to allow for the effects of turbulence and solved numerically using finite differencing. The particle energy equation, for example, is almost identical to equation (11). The equations of radiation transfer between the walls and particles are solved using a six-flux method. Gas phase reactions are assumed to be mixing-limited; thus the rate controlling step is the micro-mixing between fuel and oxidizer so that gas-phase chemical rate equations are not solved. In these cylindrical combustors, coal is fed through a central (primary) tube surrounded by a coannular (secondary) tube. The secondary air can be sent through a swirl generator prior to entering the combustor. A summary of input parameters for these simulations is given in Table 4. For the purposes of this numerical experiment, the reactors are assumed to be adiabatic. The high coal loading in the combustors allows a self-sustaining coal flame to exist (no additional combustion gases), and the swirling secondary air holds the flame near the combustor inlet. Predicted coal combustion histories using the two sets of devolatilization kinetics are shown in Figure 4 for the non-swirling (BYU) case. The change in devolatilization kinetic model results in a change of 0.25 m in the axial location of the devolatilization zone as well as a change of 1 0 % in the volatiles yield. The change in volatiles yield due to the kinetic model is also reflected in the carbon conversion at the reactor exit. In the swirling (CANMET) case, the location of the devolatilization zone is stabilized by the fluid dynamics, so that the change in devolatilization kinetics mainly influences the predicted volatiles yield (see Figure 5). Figure 5 is somewhat misleading, since 2- dimensional effects are important near the burner. This seemingly small change in predicted volatiles yield significantly changes the particle flow pattern in these simulations, as shown in Figure 6. As the particles devolatilize near the reactor inlet (bottom left in Figures 6a and 6b), the volatile gases expand, pushing the gas from the secondary stream away from the centerline. The predicted difference in gas temperature near the burner for these two cases is negligible. The high volatiles release along the centerline predicted with the Ubhayakar kinetics dampens the effects of the swirling secondary air due to the expansion of the volatiles gas, and the particles do not recirculate. In the predictions using 6 |