An Experimental Study of the Phase Separation Phenomenon and Dewatering Technique of Pyrolysis Oil
MetadataVis full innførsel
Fast pyrolysis is a promising technology for cost-efficient conversion of biomass into renewable liquid fuels. However, development of pyrolysis oils as a fuel is incomplete. Issues of phase separation and alteration of rheological properties upon storage of the pyrolysis oils lead to unfavourable fuel characteristics in handling, transportation and applications. Mitigation measures have been established for improvement such as development of suitable methods for handling pyrolysis fluids. Furthermore, pyrolysis oil quality specifications have been developed to ease commercialization of biomass-derived pyrolysis liquids for heat and power applications. Despite of establishing alleviation measures there is still a need of techniques necessary for investigations and tackling some of the issues. In this project phase separation phenomenon, modulation of rheological properties, techniques for evaluating the phase separation phenomenon and dewatering of the pyrolysis oils are investigated. The following paragraphs summarise topics explored. One of the main obstacles in using pyrolysis oils in heat and power applications is their instability upon storage, which leads to an unacceptable quality from the end user’s point of view. Due to the opaque nature of pyrolysis oils, there are presently many challenges associated with determining their stability. Thus, techniques are needed for the characterization of the phase separation of pyrolysis oils, as well as determining the underlying mechanisms of their instability. In this project the application of the Turbiscan technique for the evaluation of phase separation tendency of pyrolysis oils is presented. Investigation was done over a period of 24 hours at various temperatures and compared to the Karl Fischer method. A well stored pyrolysis oil from poplar wood and fresh pyrolysis oil from forest residue were used for investigation. For each of the oils, one batch was diluted with water in order to force phase separation, and a second batch was used without dilution. The study reveals that the Turbiscan technique makes it possible to study several aspects of phase separation in a single experiment; such as sedimentation, clarification, migration velocity and phase fraction. The advantages and potential limitation of the Turbiscan technique are discussed. Modulation of pyrolysis oil rheology upon storage was also investigated. Well-stored pyrolysis oil derived from poplar wood in a partially phase separated state was used for investigation. Homogeneous whole oils and sedimented bottom phases were examined at various thermal conditions. Results reveal that whole oils exhibit invariant low viscosity at high temperature conditions and high viscosity non-Newtonian flow at low temperature conditions. In contrast, bottom phases exhibit high viscosity non-Newtonian behaviour at all thermal conditions. Whole oils are composed of weakly linked suspended structures, whereas bottom phases are composed of a stronger network structures. The structures are shown to dissolve at high temperature conditions and degrade under shear forces. Removing water from the pyrolysis oil is a challenge because the oils are thermally unstable and consist of dissolved components in aqueous matrix. Therefore, techniques are need for removing water without altering the chemical composition in an economic feasible way. An osmotic driven membrane process Forward Osmosis (FO) which removes pure water without dissolved components and can operate at ambient temperature was investigated. The driving force for the process is the osmotic pressure difference between feed solution (FS) and osmotic agent (draw solution (DS)). In this study cellulose acetate (CA) membranes and three osmotic agents (draw solutions) were used: NaCl, MgCl2 and AlCl3 solutions; the tests were done at 25oC over 12 hours. The study demonstrates that the FO process is a promising process; water was removed reasonably well within 12 hours (~20% energy increased and 24% water reduced). This investigation shows that the NaCl solution is relevant over other solutions used; however, diffusion of the solutes from the DS to FS is a challenge given that alkali metal ions above the accepted level impair the oil. Therefore, membranes with high selectivity and draw solutions that produce high water fluxes with minimal reverse solute fluxes are needed.