Removal of H₂S, NH₃, and tars from the producer gas of biomass gasification by secondary measures.

The combination of biomass gasification and Fischer-Tropsch (FT) synthesis is a promising technology for converting biomass to “green” liquid fuel. However, the clean-up process of removing tars, H2S, NH3, and the other contaminants from the producer gas of biomass gasification is required because these contaminants are problematic in the downstream processing and some are poisonous to the catalysts used in FT process. A number of methods for the gas cleaning have been reported but these methods either have negative impacts on the environment or are too costly. This study aims to develop effective and low-cost methods for gas cleaning through experimental studies and fundamental analysis. This thesis describes studies on the removal of H2S, NH3, and tars from the producer gas of biomass gasification on a dual fluidised bed (DFB) gasifier via secondary measures with a target of the cleaned producer gas meeting the quality requirements of FT process. The key part of the thesis is the study of H2S adsorption and NH3 decomposition from the simulated producer gas by using titanomagnetite as a sorbent/catalyst in a fluidised bed quartz reactor. In addition, studies were also conducted on the modification of an oil scrubber system and preliminary results are also reported on tar removal from the actual producer gas of biomass gasification using the modified oil scrubber system. Studies on the H2S and NH3 removal are divided into two sections. The first section presents the results of H2S adsorption from the simulated producer gas by titanomagnetite in the fluidised bed quartz reactor. The aims of this section were to determine the effectiveness of titanomagnetite and the most effective operation temperatures for H2S adsorption, and to examine the effect of steam and CO in the producer gas. In the experiments, H2S concentration in the gases was controlled at 240±20 ppmv and the test gases were, respectively, (1) Ar gas, (2) simulated producer gas from 5 biomass steam gasification (CO, CO2, CH4 and H2), (3) mixture of Ar and steam, and 4) mixture of Ar and CO. The unprocessed and H2-reduced titanomagnetites were used as sorbents and operation temperatures were varied from 350 to 750C. Results from the experiments show that both of the unprocessed and the H2-reduced titanomagnetites were effective to remove H2S in Ar gas at 600°C. However, at the same temperature, the H2S removal efficiency was reduced in the simulated gas whereas the unprocessed titanomagnetite was more effective than the H2-reduced titanomagnetite. Therefore, the unprocessed titanomagnetite was further investigated to find the effect of operation temperature on H2S adsorption, and the results show that the most effective H2S removal can be achieved at 400 - 450C. It was also observed that both steam and CO in the gas mixture reduced the removal efficiency significantly although steam in the gas had more significant impacts. For NH3 removal, the studies include four parts: a). investigation of performances of both the unprocessed and the H2-reduced titanomagnetites for NH3 decomposition in Ar gas; b). investigation of the NH3 decomposition from the simulated producer gas by the H2-reduced titanomagnetite at different temperatures; c). examination of the effect of H2S presence (230 ppmv) in the simulated producer gas on NH3 decomposition using the H2-reduced titanomagnetite; d) and analysis of side reactions in the gas cleaning process and examination of the effects of gas species on the NH3 decomposition. In the last part, six gas mixtures were tested which include 1). H2 in Ar; 2). CO in Ar; 3). CO2 in Ar; 4). CH4 in Ar; 5). H2, CO and Ar; and 6). H2, CO, CO2 and Ar. All the experiments were conducted on the fluidised bed quartz reactor. In the test gases, the NH3 concentration was 2300±200 ppmv and the operating temperatures were varied from 500 to 850°C. The results from the NH3 removal experiments show that in the control Ar gas, the H2-reduced titanomagnetite had much higher activity than the unprocessed titanomagnetite to decompose NH3 6 at all of the temperatures tested, and the efficiency of NH3 removal increased with reaction temperature. The NH3 decomposition by the reduced titanomagnetite was 97.8% at 500°C, 99.7% at 600°C, and 100% at 750 and 850°C, in comparison with corresponding values of 31.6%, 34.0%, 83.9% and 93.2% for the unprocessed titanomagnetite. Therefore, the H2-reduced titanomagnetite was then employed to remove NH3 in the simulated producer gas in which the NH3 decomposition was 28.4±3.4%, 94.7±2.8% and 98.4±0.4%, respectively, at 500, 750 and 850°C. During NH3 decomposition in the simulated producer gas, side reactions have been identified and analysed which played different roles at different temperatures. Side reactions including the reverse water-gas shift reaction, the (reverse) Boudouard reaction, the (reverse) carbon Methanation reaction and the iron oxidization reaction were involved. At 500°C, the carbon formation from the Boudouard reaction significantly suppressed the activity of the reduced titanomagnetite for NH3 decomposition. In addition, it was found that 230 ppmv H2S in the simulated producer gas had a significant adverse effect on the H2-reduced titanomagnetite for NH3 removal. Negative effects by gas species in the simulated producer gas on NH3 decomposition were attributed to equilibrium reduction by H2, carbon deposition from the Boudouard reaction and carbide formation by CO, and α-Fe phase oxidization by CO2. However, CH4 was found to have only slight effect on NH3 decomposition. H2 in the simulated producer gas was also found to promote carbon formation by the reverse water gas reaction. In the meantime, H2 also had favourable effects of protecting α-Fe phase on the catalyst from oxidizing by CO2 and hindering carbon formation from CH4 decomposition. Furthermore, CO2 also had the positive effect of inhibiting carbon formation from the Boudouard reaction. In the last part of the study, experiments were conducted on a modified oil scrubber to remove tars from the actual producer gas from the DFB gasifier. The results show that without accounting for 7 an unknown compound, the oil scrubber with either biodiesel or canola oil had the efficiency of removing ~96% GC-detectable tars from the producer gas of biomass gasification. It also illustrates that the oil scrubber with biodiesel or canola oil had the ability to remove the particles, water, and GC-undetectable tars in the producer gas of biomass gasification. However, an unknown compound was detected which concentration in the outlet gas was found to be higher than that in the inlet gas. Further studies will be conducted to identify this compound and to understand the reasons for this unexpected increase in its concentration in the producer gas