Steam gasification of rice husk pellets in a dual fluidized-bed gasifier for hydrogen-rich producer gas

Abstract

Experiments were conducted in this study on steam gasification of rice husk pellets (RHP) at a 100 kWth dual fluidized bed (DFB) gasifier, which consists of a circulating fluidized bed (CFB) column and a bubbling fluidized bed (BFB) column. Effect of operation conditions including gasification temperature, steam to biomass (S/B) ratio and catalytic bed material on yield and composition of producer gas (gas product) and condensable gas compounds (tars) has been investigated in this study. Prior to running the hot experiment at high temperatures, cold runs on the gasifier system were performed to experimentally examine the behaviour of bed material fluidisation and to compare the results with those obtained from theoretical calculations.

The experimental results from the cold runs, in which the operation temperature was assumed to be 25oC, show that appreciable circulation was obtained for both silica and olivine, which had a mean particle size of 227 μm and 256 μm, respectively. While the theoretical terminal velocity (ut) was 2.07 m s-1, the required superficial gas velocity (usf) was found to be 3.43 m s-1 to achieve fast fluidisation of silica with air as the fluidisation agent in the CFB column. In the BFB column, meanwhile, the fluidisation behaviour was found to be in the bubbling fluidisation regime, obtained when the air velocity (usf) was 0.12 m s-1, falling in between the calculated umf and ut. The theoretical terminal gas velocity of olivine sand (ut) was 2.40 m s-1, which was slightly higher than that of silica sand. In order to achieve the desirable fluidisation in both CFB and BFB columns using olivine sand as the bed material, the air superficial velocities (usf) in the two columns were found to 3.77 m s-1 and 0.12 m s-1, respectively.

In hot test runs with RHP, gasification temperature was the first parameter to be investigated which was found to significantly influence the composition and yield of the producer gas and tars. With increase in temperature from 650 to 800oC, both yield and quality of the producer gas were improved with significant increase in producer gas LHV largely due to the increase in H2 concentration in the producer gas. Over the range of temperatures tested from 650 to 800°, the highest LHV of 13.06 MJ Nm-3 was achieved at 750oC, where the producer gas yield was 0.53 Nm3 kg-1od and the concentrations of H2, CO, CO2 and CH4 in the produce gas were, respectively, 24.44 vol. % , 35.34 vol.%, 20.27 vol.% and 11.92 vol.%. At this gasification temperature, tar yield and its concentration in the producer gas were also reduced. Under the optimum gasification temperature of 750oC and S/B ratio of 0.7, the total yield of GC-detectable tars was 0.53 g kg-1od, with a concentration of 9.23 g Nm-3. The tar concentration was 9.42% lower than that measured at 650oC, but slightly higher than that obtained at 800oC (7.53 g Nm-3). Over the temperature range examined, class 2 and class 3 tars was found to be dominant with phenol and benzene becoming the majority for each tar class. The heavy tar compounds in class 4 and class 5 tars were found to be less influenced with the temperature increase, leading to minimal change in their overall tar concentration.

For the second part of experiments, S/B ratio was varied by decreasing the feed rate of biomass (RHP) from 17.5 to 10.7 kgod h-1 while holding the rate of steam injection constant at 10.5 kg h-1. The gasification temperature was controlled at 750oC. The experimental results show inconsistent trend of producer gas yield and composition when the S/B ratio was increased from 0.7 to 1.1 with 0.1 ratio increments. The highest gas yield (0.57 Nm3 kg-1 od) was detected at S/B ratio of 1.0, in which CO was the most dominant gas species with concentration of 33.89%, followed by H2 (25.10%), CO2 (21.07%) and CH4 (11.94%), while the other 7.99% was made up of C2H4, C2H6 and N2. The gas LHV was 13.08 MJ Nm-3 in this case. Furthermore, at an S/B ratio of 0.8 the lowest yield of gas (0.44 Nm3 kg-1 od) was observed where the concentrations of H2, CO, CO2, CH4, C2H4, C2H6 and N2 were measured to be 26.61%, 32.57%, 21.43%, 11.82%, 4.42%, 0.70% and 2.29%, respectively, with gas LHV of 12.90 MJ Nm-3.

Further experiments were conducted with higher biomass feed rate of 14.5 kgod h-1 whereas the steam injection rate was varied from 19.2 to 18.4 kg h-1, giving the range of S/B ratio from 0.7 to 1.3 with 0.3 increments. With a constant gasification temperature of 750oC, the S/B ratio of 0.7 demonstrated the highest gas LHV reaching 12.95 MJ Nm-3, while the gas yield was found to be 0.54 Nm3 kg-1od. In this case, a consistent trend was observed within the given range of S/B ratio, in which the H2 concentration was decreased with increasing S/B ratio, leading to the highest H2 concentration of 25.01% at S/B ratio of 0.7. Meanwhile, CO, CO2 and CH4 concentrations were 33.46%, 21.03% and 12.01%, respectively. The remaining gas species accounted for 8.4% of the producer gas.

The tar yield and tar composition were found to be influenced by the S/B ratio. In a general trend, the tar yield was observed to decrease with increase in the S/B ratio, which could be correlated with the transformation of tar classes. The lowest tar yield (3.73 g kg-1 od) was obtained at S/B ratio of 0.8 where class 2 and class 3 tars were found to be the majority components and these together accounted for 82.5% of the overall tar existing in the producer gas.

Olivine sand was found to be a catalytically active bed material, which improved the producer gas yield and quality in gasification of RHP. At the optimum operation condition of operation temperature of 750oC and S/B ratio of 0.7, producer gas yield was 0.60 Nm3 kg-1 od producer gas with application of olivine sand, which was about 10% higher than that obtained with silica sand (0.54 Nm3 kg-1od) at the same operation condition. The olivine sand also favoured the water-gas shift reaction, enhancing the formation of H2 and CO2. As a result, less CO and CH4 were formed, having concentrations of 32.08% and 11.50%, respectively, with olivine; while the corresponding values with silica sand were 35.34% for CO and 11.92% for CH4. However, the gas LHV was slightly lower with the olivine sand (12.72 MJ Nm-3) than that produced with silica (13.06 MJ Nm-3).

Lower tar yield and concentration could also be seen from the use of olivine, showing its catalytic effect on the tar conversion. Apart from that of class 2 tars, the concentrations of other classes showed significant reduction with the overall tar concentration (6.15 g Nm-3) being much lower than that(9.96 g Nm-3) with silica.

To sum up, this study has successfully examined the impacts of important parameters in steam gasification of RHP. The experimental results are expected to provide important information for operation optimisation steam gasification of RHP to produce a gas product (producer gas) which is used as a fuel for internal combustion engine.