Effect of pressure on the characteristics of biomass pyroysis

Research Doctorate - Doctor of Philosophy (PhD)Pressure is one of the less studied parameters in respect to its effect on the pyrolysis of biomass. Most of the understanding of this effect is qualitative in nature. Very little insight has been reported in the literature, into the intrinsic physical and chemical processes governing the effect of pressure on the pyrolytic behaviour of biomass and its main structural components (cellulose, hemicellulose and lignin). The lack of a well scrutinised mechanistic model, able to account for the effect of pressure during biomass pyrolysis, reflects the present state of art in the field. This has hindered the development of suitable biomass thermal-converter unit operations functioning under optimised pressure, temperature and residence time, able to produce desirable products. These outputs may include the attainment of a specific phase partition (char, bio-oil and gases) and manipulation of the composition of these phases. Aiming to fill this important gap in knowledge, the primary objective of the present research was to investigate the effect of pressure and vapour-phase residence time on the pyrolysis characteristics of biomass. In pursuing this objective, we have designed and assembled a flexible fixed-bed high pressure experimental rig. Subsequently, we have deployed the rig to perform experiments on wood chips of biomass samples of eucalypt (Eucalyptus acmenoides) sapwood and heartwood, camphor (Cinnamomum camphora) heartwood, pine (Pine radiata) sapwood, and sugar cane bagasse (Saccharum officinanum). We have also undertaken experiments on commercial and experimentally extracted components of biomass (cellulose, hemicelluloses, and lignin). Our experiments comprised a thermal treatment which included a heating rate of 5 °C/min and Tmax of 450 °C, under an inert purge gas atmosphere of N2. We have executed parameterised experiments varying pressure (from 0.1 to 1.0 MPa) and vapour-phase residence time (0.6, 1.5, and 15 min). We have also carried out experiments on a high-pressure thermo-gravimetric apparatus, under operating conditions similar to those deployed in the experimental rig. The analytical instrumentation consisted of a micro gas chromatograph, allowing on-line analysis and quantification of the gaseous species, and a gas chromatograph-mass spectrometer to perform qualitative and quantitative analysis of the tarry fraction. Proximate and ultimate analyses of samples and pyrolysis products were performed using standard analytical techniques. From this set of experiments, we were able to demonstrate a dependence of product phase yields (char, tar and gases), their composition, and their quality on the structural composition of woody biomass. We have found that lignin contributes the most, and cellulose the least to the formation of char. We have also demonstrated that high pressures and long vapour-phase residence times act concomitantly to produce a high yield of char and gases, as a consequence of a reduction in the yield of tar. Cellulose responded to increasing pressure, by enlarging the yield of char, as a result of secondary decomposition of levoglucosan. Levoglucosan forms during the thermal decomposition of cellulose and, at high temperature and under mass transport limitation (high pressure, and high residence time, for instance), easily decomposes to secondary char and gases. We then concluded that, from a technological perspective, biomass materials of large lignin content submitted to high pressure and elevated vapour phase residence timeoffer promising potential for production of high yield and high quality char. We have demonstrated that high pressure and long residence times act by altering the overall gas and tar compositions, as well as the char quality. In regard to the gas phase, we have concluded that, high pressure and long residence times promote the formation of CO2 (mainly from the carbohydrate fraction) in comparison to the formation of CH4 and CO. With respect to the condensed phase, we have established that Class I compounds in wood tar (cellulose and hemicellulose-derived compounds) experienced a drastic reduction in their yield at elevated pressure and prolonged residence time. In contrast, species comprising Class III (aromatic and polycyclic compounds) increased their abundances. The char quality, represented by the content of fixed carbon, rose at elevated pressures. We have also demonstrated experimentally that, biomass materials characterised by high lignin content tend to produce higher yields of primary char, which contains a relatively high percentage of fixed carbon. We have developed of a novel Arrhenius-type expression describing the rate of formation of secondary char and gases, based on a microscopic mass balance of the species produced in the pyrolysis of cellulose. The model is based on an extended Broido-Shafizadeh semi-global mechanistic approach that incorporates the secondary decomposition of vapour tar into secondary char and gases. The model accounts for the interplay between the kinetics and mass transfer during the pyrolysis of cellulose at high pressures. We have applied a similar approach, based on the Koufopanos et al. semi-global mechanism, to the remaining two structural components of biomass (hemicelluloses and lignin). The kinetic rate parameters (and the mass transport coefficient) for the extended stage 3, the thermal decomposition of vapour tar into secondary char and gases, were then determined for all three biomass structural components. Our result constitutes one of few attempts to describe quantitatively the effect of pressure on the decomposition of cellulose, hemicelluloses and lignin, contributing to progress in the field. Based on the observation that decomposition of individual components of biomass is well described by individual models, we have succeeded to develop a kinetic model that accounts for the thermal decomposition of the entire woody biomass. Relying on the evidence gathered from the experiments carried out both on our experimental rig and on the high-pressure thermogravimetric apparatus, we have concluded that the superposition principle satisfactorily describes the pyrolysis of biomass, both at low and high pressure. This salient conclusion underpinned the development of a complete pressure-dependent kinetic model for the pyrolysis of woody biomass that requires as inputs only the biomass composition, in terms of cellulose, hemicelluloses and lignin content. The model fits well experimental measurements for hardwood and softwood biomass species, confirming the validity of the superposition principle