Abstract
Formation of sticky layers on bed particles has been considered as a prerequisite for bed agglomeration in fluidized bed combustion of wood-derived fuels. The present investigation was undertaken to determine the quartz bed particle layer formation process in fluidized bed combustion of wood-derived fuels. Bed material samples from three different appliances, bench-scale bubbling fluidized bed, full-scale bubbling fluidized bed, and full-scale circulating fluidized bed, at different sampling times from startup with a fresh bed were collected. Scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDS) and X-ray diffraction (XRD) were used to explore layer morphology and chemical composition and to gain information on crystalline phases of the layers and coatings. Significant differences in layer morphology and composition were found for quartz bed particles with different ages. For bed samples with operational duration of less than 1 day, only one thin Ca-, Si-, O-, and K-rich homogeneous quartz bed particle layer that has a relatively high K/Ca molar ratio was found. For quartz bed particles with an age from around 1 day to 2 weeks, an outer more particle-rich coating layer was also found. During the initial days of this period, the layer growth rate was high but decreased over time, and decreasing K/Ca and increasing Ca/Si molar ratios in the inner bed particle layer were observed. For bed particles with age between 2 and 3 weeks, a much lower layer growth rate was observed. At the same time, the Ca/Si molar ratio reached high values and the K concentration remained on a very low level. In addition to these layer formation processes mentioned, also an inner–inner/crack layer was also formed in the circulating fluidized bed quartz bed particles simultaneously with the inner bed particle layer.
Highlights
Biomass fuels, provided for potentially renewable and CO2neutral energy sources, have been used widely during recent decades
The bed agglomeration mechanisms during Fluidized bed combustion (FBC) of biomass fuels in a quartz bed were summarized by Brus et al.,[9,10] De Geyter et al.,[11] and Grimm et al.,[2] who concluded that the agglomeration proceeds with an initial bed particle layer formation, followed by subsequent viscous flow sintering and agglomeration for fuels rich in Ca and K, i.e., wood-derived fuels, or direct adhesion of bed particles by partly molten ash-derived silicates or phosphates for fuels rich in Si and/or P, e.g., straws and agricultural residues
The element Si was excluded from the analyses shown in panels a and b of Figure 5, i.e., for samples taken from Bubbling Fluidized Bed (BFB) at 5 kW to avoid possible influence from other parts than the thin layer because of limited spatial resolution
Summary
Biomass fuels, provided for potentially renewable and CO2neutral energy sources, have been used widely during recent decades. Fluidized bed combustion (FBC) is considered as one of the most suitable technologies for heat and power production, because of high fuel flexibility, low process temperature, and emission control. Bed agglomeration could be one potential operational problem, which may lead to a non-scheduled shutdown of the combustion plant in severe cases with significant economic losses. Frequent bed change is a potential measure for reducing the bed agglomeration risk for some fuels, it definitely brings about additional cost and is not economically sustainable on a longterm basis. Any predictive methodologies for assessing the agglomeration tendency are important for operated reactors. Multiple layers with different properties and compositions have often been found around bed particles during combustion of biomass.[9,12−14] It appears that the inner layer compositions depend upon the composition of both the bed and the fuel, whereas the outer layers are more related to the fuel ash compositions.[7,9,11,14,15]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
