Interaction between SO 2 from flue gas and sorbent particles in dry FGD processes

Abstract

Among the technologies to control SO 2 emission from coal-fired boilers, the dry flue gas desulphurization (FGD) method, with appropriate modifications, has been identified as a candidate for realizing high SO 2 removal efficiency to meet both technical and economic requirements, and for making the best quality byproduct gypsum as a useful additive for improving alkali soil. Among the possible modifications two major factors have been selected for study: (1) favorable chemical reaction kinetics at elevated temperatures and the sorbent characteristics; (2) enhanced diffusion of SO 2 to the surface and within the pores of sorbent particles that are closely related to gas-solid two-phase flow patterns caused by flue gas and sorbent particles in the reactor. To achieve an ideal pore structure, a sorbent was prepared through hydration reaction by mixing lime and fly ash collected from bag house of power plants to form a slurry, which was first dewatered and then dried. The dry sorbent was found capable of rapid conversion of 70% of its calcium content at 700 oC, reaching a desulphurization efficiency of over 90% at a Ca/S ratio of 1.3. Experiments confirmed that the diffusion effect of SO 2 is an important factor and that gas-solid two-phase flow plays a key role to mixing and contact between SO 2 and sorbent particles. For designing the FDG reactor, a new theoretical drag model was developed by combination of CFD with the Energy Minimization Multi-Scale (EMMS) theory for dense fluidization systems. This new drag model was first verified by comparing calculated and measured drag values, and was then implemented in simulation of gas-solid two-phase flow in two circulating fluidized beds with different sizes and flow parameters. One riser has diameter and height of 0.15m×3m and another one 0.2m×14.2m. Their superficial gas velocities are 4 and 5.2m·s −1, respectively, and the circulating rate 53 and 489 kg·(m −2·s −1). FCC particles were used in both cases. The results show that not only the static pressure drop along the riser height, but also radial distributions of particle volume fraction have been very well predicted in comparison with experiments. The new drag model is expected to shed more light on the further improvement of SO 2 diffusion to solid sorbent and optimization of reactor structure.