Compared with the typical solid fuels (i.e., pulverized coal, biomass powder, etc.), the black liquor combustion performs the unique characteristics of the high swelling ratio, unsteady flame, and high corrosivity. During this combustion process, to realize the efficient recovery of the sulfides, it is well-established that a stable reductive zone at the bottom of the furnace is necessary, that is, the char bed plays an important role in enhancing the recovery for the Kraft recovery boiler. On the one hand, it protects sulfide in the molten smelt from reoxidation to sulfate. On the other hand, it provides an interface for air/carbon contact To deeply understand the combustion characteristics of black liquor to prevent the corrosion, slagging and fouling during the Kraft recovery boiler operation, the computational fluid dynamics (CFD) approach is adopted due to its advantages of the high efficient and low cost, in which the complex combustion processes in the furnace can be obtained and the key information for optimizing the furnace design is provided.
In this paper, a computational fluid dynamic (CFD) approach modified by the user defined functions (UDFs) that is programed by C language is developed to predict black liquor combustion process (i.e., drying, devolatilization and char burning) in a Kraft recovery boiler. For example, DEFINE_DPM_SOURCE macro is defined to calculate particle volatiles release rate. DEFINE_DPM_SWITCH macro is defind for the stage transition rules of the combustion stage. DEFINE_PR_RATE macro is adopted to redefine the reaction rate The material and energy transformations of the continuous phase in each stage can be defined by to characterize the source term variations. For transitions, it requires the use of For the particle surface reaction rate during the char burning stage, it is necessary to by the. In addition, the combustion behavior of particles on the char bed can be defined using the DEFINE_DPM_BC macro. Further, they are embedded into FLUENT software to predict the variation of the black liquor particle diameter during the combustion process, the predicted result is nearly consistent with that of the experiment.
The results show that, the modified CFD model is applied to the combustion simulation of a Kraft recovery boiler, and the relative deviations between the simulated and the measured values of key parameters are less than 5%. Moreover, two different distributed modes (i.e., cross-counter, tangential) for secondary air are adopted to further suggest an improved choice for this Kraft recovery boiler. Compared with the tangential arrangement, the cross-counter arrangement mode is favor to create the swirl flow with a smaller radius in the furnace, resulting in more uniform temperature and species distribution, demonstrating more obvious advantages in reducing coking and slagging.
 

CFD simulation,Black liquor combustion,Kraft recovery boiler,UDFs,Secondary air arrangements