BIOMASS CHARACTERIZATION AND GASIFICATION IN A FLUIDIZED BED

Abstract

The exploitation of renewable energy sources for meeting the ever increasing energy requirements has become more demanding for those developing countries which are importing significant amounts of crude petroleum, natural gas and petroleum products. Among the several alternative renewable energy sources, biomass occupies a unique place. Biomass materials like agricultural, agro-processing and forestry residues are available in abundance in India. Some of them pose disposal problems due to their low fodder and fertilizer values, e.g. rice husk, bagasse, sawdust and pressmud (a waste from sugar industry), etc. Thermal gasification for energy utilization of such waste biomass materials is an attractive proposition. The well developed moving bed gasifiers have not been found suitable for such materials. A fluidized bed gasifier accepts the biomass of low bulk density and high ash content such as agricultural/agro-industrial and forestry wastes. The present experimental study was undertaken to understand the fluidization behaviour, thermal degradation behaviour and kinetics of five most commonly and abundantly available agricultural, agro-industrial and forestry residues in India. These included village rice husk, mill rice husk, sawdust, bagasse and pressmud. Further, in the study, the village rice husk and sawdust were employed for studying their gasification characteristics in a fluidized bed gasifier. Experimental Programme Fluidization Experiments The fluidization behaviour of the five biomass materials as binary mixtures with inert carrier solids like sand and bauxite were studied under cold flow atmosphere conditions in a set-up comprising of a vertical plexiglass fluidizing column of 76 mm id IV and 1000 mm high and fitted with a multiorifice type distributor plate. Experiments were conducted for different weight percent of biomass content in the bed mixture (charge). Thermal Degradation Behaviour and Kinetics The biomass materials were subjected to thermal analysis in a Stanton Redcraft thermo analyzer. Two heating rates (20°Cmin'' and 40°Cmin"') and two experimental atmospheres (nitrogen and air) were used to obtain TGA, DTG and DTA thermograms. Thermal degradation characteristics and the kinetic parameters namely, activation energy, pre-exponential factor and order of reaction were obtained from these thermograms. Available integral and differential methods like these proposed by Coats and Redfern, Agarwal and Sivasubramanian, Horowitz and Metzger, Freeman and Carroll, Reich and Stivala and Piloyan and Novikova were used to determine kinetic parameters. Fluidized Bed Gasification of Village Rice Husk and Sawdust Fluidized bed gasification of village rice husk and sawdust were conducted in a 100 mm id, 1650 mm total height fluidized bed gasifier unit developed in the laboratory. The unit comprised of the fluidized bed reactor, the biomass feeding unit, high efficiency cyclone and an after burner. A multiorifice type distributor plate was used to support the fluidizing solids and to distribute the fluidizing and gasifying air. The gasification experiments were conducted at four fluidization velocities (0.53, 0.59, 0.68 and 0.73 ms'1) and at five equivalence ratios (0.20, 0.25, 0.30, 0.35 and 0.40). Gas and tar sampling trains as per latest procedures were used. The gas was compositionally analysed using a gas chromatograph. The tar content of the raw gas was analysed gravimetrically. The performance of the gasifier under different operating conditions of fluidization velocity and equivalence ratio was evaluated in terms of temperature profile in the reactor, gas composition, gas heating value, gas yield, carbon conversion efficiency, tar and particulate content of the gas and the thermal efficiency of the gasifier for the two feedstocks. Results and Discussion Preliminary Experiments In the preliminary experiments, the properties of the five biomass were determined to assess their suitability and feasibility for use as fuels/feedstocks for gasification. Their low bulk density values (< 200 kgm"3) indicated that they were unsuitable as feedstocks for a moving bed gasifier. A fluidized bed gasifier accepts them as feedstocks. All the five biomass fuels had a low nitrogen content and a sufficiently good heating value. Fluidization Behaviour of Biomass Materials Attempts to fluidize a biomass material alone as a single component bed charge indicated that a low bulk density (bulk density < 200 kgm"3), complex shaped biomass material was not fluidizable easily and that an inert solid (carrier solid) must be added to the bed charge to make a binary mixture to facilitate its fluidization. Significant hysteresis was observed in the plots of pressure drop versus superficial air velocity (increasing and decreasing mode). Fluidization of binary mixtures of carrier solid (sand or bauxite bed with H/D=l) and 1 to 9 wt% of each biomass material indicated good fluidization of the bed charge upto a biomass content of 8 wt%. The quality of fluidization deteriorated with each one percent increase in the biomass content of the bed charge (visually observed) and that the mixture was not fluidizable beyond 8wt% of the biomass content in the bed. It was also observed that with increase in the biomass content in the bed mixture, the value of minimum fluidization velocity also increased. Thermal Degradation Characteristics of Biomass Materials TGA, DTG and DTA curves showed the several distinct zones during which the thermal degradation process of the biomass. A natural break in the slope of the vi TGA curve is indicative of a change in the rate of weight loss due to drying, devolatilization and degradation of biomass components namely hemicellulose, cellulose and lignin. The nature of TGA, DTG and DTA curves clearly indicated the number of reaction (degradation) zones for the thermal degradation of the five biomass materials under different experimental conditions. From the thermograms, the rates of degradation and weight of residue at the end of the degradation could be determined. For village rice husk, four reaction zones were identified, while only three reaction zones could be seen for mill rice husk. During pyrolysis, the maximum rates of degradation for village rice husk were 11.7 and 23.1 wt% min"1 at heating rates of 20 and 40°Cmin"1 respectively. The thermal degradation in air showed only two distinct temperature zones for both village rice husk and mill rice husk with 14.9-17.5 wt%min"' at 20°Cmin"' and 37-39.51 wt%min"' at 40°Cmin"1 heating rates, in the first reaction zone. The degradation rates in the second reaction zones were lower. For sawdust, a five clear temperature (reaction) zones could be identified. The maximum rates of degradation were 29.7 and 56.9 wt%min'1 at 20 and 40°Cmin'' heating rates respectively. The residues were 17.5% at 20°Cmin"1 and 25.0% at 40°Cmin"1. During thermal degradation in air, four distinct degradation zones, other than the one of drying, were identified. Maximum rates of degradation were during second reaction zone with the values of 46.6 wt% min'1 and 51.6 wt%min"1 for 20 and 40°Cmin" heating rates respectively. The thermal degradation of bagasse in nitrogen indicated three reaction zones with the second zone having the maximum degradation rates of 22.5 wt%min"1 and 49.6 wt%min'' for 20 and 40°Cmin'1 heating rates respectively. Air degradation had five reaction zones. Pressmud, a sugar mill waste, showed distinct degradation behaviour in comparison to other biomass materials. Pyrolysis behaviour showed poor devolatilization with maximum degradation rates of 9.7 wt%min'' and 14.0 wt%min"' at 25 and 40°Cmin"1 respectively. However in an oxidizing environment rate of vn was as high as 90 wt%min'1 at lOXmin"1 which showed that pressmud has the potential to be considered as a feedstock for gasification. Kinetics of Thermal Degradation TGA curves were used to determine kinetics of thermal degradation of the biomass materials in nitrogen and air atmospheres using differential and integral methods. In all six methods as proposed by Coats and Redfern (1964), Agarwal and Sivasubramanian (1987), Horowitz and Metzger (1963), Piloyan and Novikova (1967), Reich and Stivala (1982) and Freeman and Carroll (1958) were used to determine kinetic parameters assuming a simple, single-step, irreversible reaction. The entire degradation zone was also represented by a single-step, irreversible reaction using Agarwal and Sivasubramanian approximation for the integral analysis of kinetics. For all the biomass materials, at the two heating rates and under nitrogen and air atmospheres, the kinetic parameters were determined by using a least-square best-fit approach. For the integral method of Coats and Redfern, and Agarwal and Sivasubramanian, the n was varied from zero to 3.0 to obtain the best-fit linear relation to determine the activation energy E, and the pre-exponential factor, ko. For any n, the value of E obtained by Coats and Redfern method and Agarwal and Sivasubramanian method were the same. The pre-exponential factor, however, differed slightly. From the results it was clear that the entire range of degradation could be represented by a singlestep, irreversible reaction using Agarwal and Sivasubramanian (1987) approximation for integration, with reasonable confidence for all the biomass materials. It was found that as the value of n increased, the activation energy also increased for the best fit correlation using integral method. Therefore, it is not the activation energy alone, but the value of order of reaction n which gave the best fit for the TGA data, alongwith E and k<j should be taken for the design of pyrolysis and gasification reactor. The kinetic parameters indicated that the best fit kinetic parameters varied in the range 0.50 < n <2.50, 27.84<E < 132.9 kJmol"1 and 3.21 x 101<ko<1.414xl012 min"1 for all the biomass materials. For air degradation, the values of E are found to be viii higher than those for pyrolysis at any value of n. The values of n, E and k0 are found to be within the range reported by other investigators. Fluidized Bed Gasification of Village Rice Husk and Sawdust The performance of the gasifier-unit, as a whole, was found to be satisfactory for both village rice husk and sawdust gasification. The average bed temperature varied between 575 to 782°C for village rice husk gasification and 680 to 791°C for sawdust gasification. The average free board temperatures were found to be lower than the average bed temperatures by 79 to 131°C for village rice husk and 81 to 92°C for sawdust. This was because of the absence of any external heat source to assist the gasification process, as has been employed by other investigators. Gasifier temperatures increased with increase in the equivalence ratio and/or with increase in the fluidization velocity. As the equivalence ratio increased, the concentration of C02 increased and the concentrations of fuel gases viz., CO, H2, CH4 and C2Hm decreased. Increase in fluidization velocity lead to the similar effect. Among the fuel gases, CO had the highest concentration (11.2-19.4% vol. for VRH; 12.8-19.9% vol. For SD) followed by H2 (2.7-2.8% vol. for VRH; 3.1-4.3% vol. for SD), and CH4 (1.5-2.3%vol. for VRH; 2.0 - 2.7% vol. for SD). The gas higher heating value (HHV) ranged from 2.35 to 3.85 MJNm' (gas at normal temperature and pressure) for village rice husk (VRH) and 2.81 to 4.14 MJNm'3 for sawdust (SD). The higher heating value of the product gas from both village rice husk and sawdust decreased linearly with increase in the equivalence ratio and fluidization velocity. The gas yield varied from 1.8 to 2.14 Nm3 kg"1 DAF for village rice husk and 1.88 to 2.13 nm3 kg"1 DAF for sawdust. In both cases, the gas yield increased with equivalence ratio (ER) upto ER=0.35 and thereafter it decreased. The energy output (the energy value ofthe gas per kg ofVRH or SD) were found to be in the range of 4.7 to 7.3 MJkg"1 DAF for VRH and 5.6 to 8.1 MJkg" DAF for SD. Carbon conversion efficiency varied from 52.03 to 65.94% for VRH and 57.18 to 66.93% for SD. IX The gravimetric method allowed the determination of high boiling tar compounds only. The tar content of the raw product gas slightly decreased with increase in equivalence ratio for both village rice husk and sawdust. Fluidization velocity did not seem to affect the tar content. Particulate matter content of the gas did not exhibit any specific trend with either equivalence ratio or fluidization velocity. The gasifier thermal efficiency ranged from 51.43 to 72.30% for sawdust gasification and 36.39 to 57.04% for village rice husk gasification.