Activity of water in pyrolysis oil—Experiments and modelling
ILLE, YANNIK; KRÖHL, FABIAN; VELEZ, ALEXIS; FUNKE, AXEL; PEREDA, SELVA; SCHABER, KARLHEINZ; DAHMEN, NICOLAUS
JOURNAL OF ANALYTICAL AND APPLIED PYROLYSIS
ELSEVIER SCIENCE BV
Año: 2018 vol. 135 p. 260 - 270
Fast pyrolysis is an option to convert lignocellulosic biomass into a large fraction of liquid bio-oil besides some char/ash and non-condensable gas. When ash-rich material is used, this usually results in a higher water content of the pyrolysis oil and a possible phase separation. To avoid phase separation, fractional condensation systems have been developed. Choosing appropriate condensation temperatures is essential to adjust the desired water content and bio-oil composition. For the modelling of such systems, detailed knowledge about the phase equilibrium of bio-oils is necessary, especially when aerosol formation is considered. This study presents the development of a model that is capable of describing the highly non-ideal behavior of these complex biomass derived mixtures. Straw based bio-oils from the bioliq® fast pyrolysis process are investigated. The required vapor liquid equilibrium data is measured in a phase equilibrium cell. The results are used for the development of suitable surrogate mixtures by studying about 200 possibly suited components to represent the heavy fraction of the bio-oil. For the calculation of the activity coefficients the gE-model Modified UNIFAC is used. Experimental and calculated data is presented and compared, with the focus on the activity of water as a key property for the thermodynamic behavior. The measured activity of water validates its strong non-ideal behavior in bio-oil. For the bio-oil investigated in this work, and according to the predictions of UNIFAC, a pseudo component formed by 3 groups of aromatic alcohol and 12 aromatic carbons (as in 3,4,4′-biphenyltriol) shows the thermodynamic behavior that best represents the unknown fraction of bio-oil. Depending on the choice of the surrogate mixture, the results can differ more than 100%. This demonstrates the importance of a carefully chosen model as an essential tool for simulating fractional condensation systems.