1. Temperature Profile
Figure 3 shows the axial temperature profiles along the combustor height for the co-firing of the mixed EB and PNS at EF2 = 0, 0.28, 0.53, 0.77, and 1, when firing mixed fuel at EA of approximately 40%, 60%, and 80%. As can be seen in Fig. 3 (a), (b), and (c), the EA had significant effects on the average temperature in each co-firing test. By increasing EA at the fixed heat input, the volume of oxidizing air was somewhat higher; therefore, the temperature attained the lower value.
The secondary fuel energy fraction had a significant impact on the axial temperature. Across the wide range of co-firing tests, the average temperature in the bed region (at Z = 0.23−0.46 m) was in the range of 750−880°C, while it was approximately 15% lower at 650−780°C in the freeboard area. When firing pure PNS (EF2 = 1), the temperature was highest in the main combustion chamber, with its lowest value in the upper chamber. These results could be due to the high density and coarse particles of the PNS; therefore, the fuel was mainly fired at the bottom part of the combustor. Among the other fuel options, mixed fuel with a higher EB proportion had a more uniform temperature profile, due to the fixed energy release rate of this experimental procedure.
2. Local Heat Transfer Coefficient (hlocal)
Figure 4 shows the local heat transfer coefficients at r/R = 0, ± 1/3, ± 2/3, and ± 1 for the co-firing of the EB (primary fuel) and the PNS (secondary fuel) at EF2 = 0, 0.28, 0.53, 0.77, and 1 at Z = 0.46 m. In the bed splashing zone (Z = 0.46 m), the radial heat transfer coefficients at the center of the combustor and the combustor wall (r/R = 0 and ± 1, respectively) were slightly higher than those at other radial locations of the probe. This phenomenon could be due to the flow pattern of the bed material and the bubble frequency [27]. By using the air distributor’s special design, the sand bed was expanded in an upward direction with a swirling movement; therefore, the particles generally rose up in the center and dropped down near the wall. The heat transfer at the center of the combustor was enhanced by the high solid concentration’s accelerated interaction frequency (or the rate of renewal) between bed particles and the immersed surface [33]. Meanwhile, in the area close to the combustor’s wall, bubbles frequently occurred due to the strong outward centrifugal force of swirling particles. This increased the turbulence of the flow and the surface renewal frequency rate leading to a slightly higher level of heat transfer. The same behavioral trend was found in the conical FBC studied for gas-solid radial heat transfer by using two-dimensional (2D) modeling of the air sand bed [34].
Figure 5 shows the local heat transfer coefficients at r/R = 0, ± 1/3, ± 2/3, and ± 1 for the co-firing of the EB (primary fuel) and the PNS (secondary fuel) at EF2 = 0, 0.28, 0.53, 0.77, and 1 at the TS-FBC’s five different heights. In comparison with the splashing or dense bed zone, the local heat transfer in the freeboard region (Z = 2.57 m) was quite uniform owing to the minimal effects of swirling flow at a very low solid concentration. In this area, gas convection was rather dominant; therefore, the gas behavior’s effects on heat transfer were quite obvious.
3. Average Heat Transfer Coefficient at Each Combustor Height (hz, avg)
Figure 6 shows the effects of EF2 on the hz,avg in the TS-FBC at excess air (EA) levels of approximately (a) 40%, (b) 60%, and (c) 80%. The EF2 noticeably affected the hz, avg in both the dense bed and freeboard regions; however, the hz, avg in the bed region increased at higher EF2, whereas the opposite occurred in the freeboard region. Moreover, the EF2 also demonstrated a significant effect on the location where the maximum hz, avg occurred. As can be seen in the axial profiles in Fig. 6 (c), the highest value of the hz, avg for firing pure EB (at EF2 = 0) was found at the connecting pipe, while the lowest level was found at the upper part of the combustion chamber for firing pure PNS (at EF2 = 1). These occurrences could be due to the combustor design. This combustor’s connecting pipe was designed to separate coarse fuel particles from fine fuel particles. As the coarse fuel particles and the inert bed material, due to centrifugal force, fell down to the dense bed zone, a small number of fine fuel particles were expected to escape to the top combustor and fire in this area.
Due to the larger particle size and higher fuel density of the PNS, co-firing of the blended fuel with a higher PNS proportion demonstrated that the vigorous combustion occurred at the bottom part of the combustion chamber. Therefore, the maximum of hz, avg was due to the solid particle convection (sand, fuel, and ash) at the intense heat released rate. With an increasing proportion of the EB, the combustion zone drifted to the higher level as the fine fiber of the shredded EB was easy to ignite and flew to the top of the combustor. When focusing on the hydrodynamics behavior of this fuel option, the upward-swirling flow of hot gas was impeded by a reduced cross-sectional area of the connecting pipe; therefore, the pressure reduced and the hot gas velocity reached the maximum value leading to the highest value of hz, avg.
Figure 7 shows the effects of EA on the hz, avg in the TS-FBC with three different secondary fuel energy fractions at: (a) 1, (b) 0.28, (c) 0.53, and (d) 0.77. As shown in Fig. 7, EA had significant effects on the hz, avg in both dense bed and freeboard regions. When increasing EA (from 40–80%) across the wide range of EF2, the hz, avg increased by about 22−27 W/m2K (or 11%, on average) in the bed region (at Z = 0.23 m and 0.46 m), and 18−53 W/m2K (or 22%, on average) in the freeboard region (at Z = 2.08 m and 2.57 m).
When investigating increases in EA across the scope of this study, the dense bed region appears to provide better mixing and dispersal of bed, ash, and fuel particles, as well as improving particle circulation [25]; therefore, the value of the heat transfer coefficient increased. The higher-value heat transfer coefficient could also be due to low heat resistance between the particles and the heat transfer surface at a high gas velocity [35]. In the freeboard region, the heat transfer coefficients also increased in value with the increasing EA; however, this was a lower-level increase due to very dilute solid particles and the process being dominated by gas convection.
4. Average Heat Transfer Coefficient for Each Operating Condition (havg)
Figure 8 shows the average heat transfer coefficients (havg) for the various fuel options and the range of excess air (EA), with only EA showing significant effects on the havg. The havg increases with increasing EA or increasing gas inlet velocity, indicating good agreement with the modeling study of Abdelmotalib and Im [34].
As shown in Fig. 8, the havg slightly increased with increasing EF2 (higher than the PNS proportion); however, the current study could not observe the effects of fuel properties in terms of EF2 on the havg to a noticeable extent. When increasing EF2 or increasing the proportion of PNS so it had a higher LHV than the EB, the combustor load was decreased to maintain the same heat release rate of the blended fuel. Under this condition, the deviation of fuel-ash content should be considered, as it is one of the key parameters in particle heat transfer. As revealed in the literature, heat transfer between cluster particles and the sampling probe could be enhanced by the collision frequency of solid particles of fuel, ash, and inert bed material in the bed region [31]. However, the fuel properties within the scope of the current study were not found to have significant effects on the havg, with the reason possibly being the limited area of the sampling probe’s heat transfer surface.