Comparative investigation on gasification performances of co-pyrolytic char from bio-oil distillation sludge and rapeseed cake: Decomposition, kinetic, structural and prediction characteristics

Abstract

The formation of bio-oil distillation sludge (DS) severely hampers the clean and continuous operation of biomass refinery system in terms of corroding the equipment and hazarding to human health, and thus the effective degradation and further disposal of DS is in accordance with the contaminant management and cleaner production. In this study, we proposed a co-pyrolysis strategy to optimize the gasification reactivity of DS to realize rapidly and efficiently converting DS into syngas. The co-pyrolysis feedstocks were DS and rapeseed cake (RC) under the carbonization temperatures of 800 °C and 1000 °C, followed by being gasified at 850–950 °C to estimate the gasification characteristics and kinetics. The results indicated the gasification reactivity of RC char was markedly higher than that of DS char, thereby with the potential to facilitate biochar reactivity of DS. The co-pyrolytic char reactivity obtained at temperature of 800 °C was enhanced with the elevated RC ratio but was all lower than that of mono-pyrolysis char, whereas the gasification reactivity of biochar pyrolyzed in 1000 °C was continuously strengthened as the DS ratio decreased from 100 % to 0. The kinetic results also indicated the apparent activation energy were increased with adding more proportion of DS, which resulted in a higher energy barrier to initialize the gasification reaction. The carbon arrangement was a critical key factor to modify the biochar reactivities, from which the large amounts of amorphous carbons and defects were promoted the biochar reactivity of RC char, whereas the ordered arrangement of crystallites for DS char was scarcely conducive to the diffusion effects of gasifying agents in carbon atoms and further diminishing the gasification reactivity. Furthermore, machine learning (ML) algorithm was for the first time adopted to simulate the co-gasification characteristics as a function of reaction time and blend ratios, and presented a higher consistency with experimental data, indicating the potential value of ML in predicting the co-gasification behaviors. Our findings can provide a feasible strategy to recover DS waste in a clean and efficient pathway, and layer a foundation for further perfecting the biomass refinery system to match the upstream production with downstream waste processing.