Oxy-combustion and oxygen-enriched combustion appear as promising options for CO 2 capture. In this study, the effect of the recycled flue gas (RFG) ratio on lignite oxy-combustion has been experimentally studied in comparison with oxygenenriched (O 2 /N 2 ) combustion in a 30 kW th circulating fluidized bed reactor. Oxygen concentrations range from 21 to 28% for both cases, while RFG ratios vary between 0.72 and 0.63 in the oxy-combustion case. The influence of oxygen concentration in the combustion atmosphere on temperature, gaseous emissions, and carbon-based combustion efficiency were investigated by keeping the excess oxygen ratio constant at 1.45 for all cases. Temperatures in the combustor increase with increasing O 2 concentration. CO 2 emissions are much higher in O 2 /RFG combustion. CO and N 2 O emissions decrease with O 2 enrichment, while NO and SO 2 emissions are promoted for both modes. Under O 2 /RFG mode, CO, NO, N 2 O, and SO 2 mass concentrations (mg/MJ) are lower compared to O 2 /N 2 mode. Higher oxygen concentrations increase combustion efficiencies for both cases. Similar O 2 concentrations result in similar combustion efficiencies. Oxy-combustion results are considered as the most favorable option for CO 2 capture along with the competitive combustion efficiencies and less emissions per unit energy.

A 30 kWth Circulating fluidized bed (CFB) combustor is experimentally and numerically investigated under cold flow conditions. Barracuda software based on Computational Particle Fluid Dynamics (CPFD) method is utilized for simulations. The influences of bed inventory and drag model on flow hydrodynamics were investigated considering pressure and velocity profiles and particle concentration. Two advanced drag models, namely Energy minimization multi-scale (EMMS) and Wen-Yu/Ergun were selected for this study. The simulations were performed with initial bed material masses of 3.79, 4.55 and 5.20 kg corresponding to 2.5, 3 and 3.5 diameters height of riser, respectively. With increasing bed inventory pressure drops and solid concentration increase. The axial particle velocities slightly change with bed inventory. The comparison of simulation results with experimental measurements was resulted in good agreement (<5%) with both models. The simulation with EMMS drag model predicted the pressure profiles more accurately than Wen-Yu/Ergun drag model. The profiles of particle volume fraction and axial velocity demonstrate that core-annulus flow pattern was captured by both models. But EMMS drag model was better in revealing the meso-scale structures at instantaneous particle concentration distribution. Moreover, the influence of particle size distribution on particle volume fraction and particle velocity profiles is also investigated with two drag models.

No abstract