Til dokument

Sammendrag

Efficient digestate dewatering is crucial to reduce the volume and transportation cost of solid residues from anaerobic digestion (AD) plants. Large variations in dewatered cake solids have been reported and predictive models are therefore important in design and operation of such plants. However, current predictive models lack validation across several digestion substrates, pre-treatments and full-scale plants. In this study, we showed that thermogravimetric analysis is a reliable prediction model for dewatered cake solids using digestates from 15 commercial full-scale plants. The tested digestates originated from different substrates, with and without the pre-AD thermal hydrolysis process (THP). Moreover, a novel combined physicochemical parameter (C/N•ash) characterizing different digestate blends was identified by multiplying the C/N ratio with ash content of the dried solids. Using samples from 22 full-scale wastewater, food waste and co-waste plants, a linear relationship was found between C/N•ash and predicted cake solids for digestates with and without pre-AD THP. Pre-AD THP improved predicted cake solids by increasing the amount of free water. However, solids characteristics like C/N ratio and ash content had a more profound influence on the predicted cake solids than pre-AD THP and type of dewatering device. Finally, C/N•ash was shown to have a linear relationship to cake solids and reported polymer dose from eight full-scale pre-AD THP plants. In conclusion, we identified the novel parameter C/N•ash which can be used to predict dewatered cake solids regardless of dewatering device and sludge origin.

Til dokument

Sammendrag

Earlywood samples of unmodified and acetylated radiata pine were exposed to the brown-rot fungus Rhodonia placenta for 1, 2, 3 and 4 weeks for unmodified samples and 10, 16, 24 and 28 weeks for acetylated samples. Longer incubation periods were used for acetylated samples based on the hypothesis that given enough time under favourable conditions the fungus would eventually degrade the wood. After exposure, samples were weighed and chemically characterized by ATR-FTIR analysis, acetyl content by saponification, and hydroxyl (OH) accessibility by deuterium exchange. Longer incubation times for acetylated samples led to comparable levels of mass loss between unmodified and acetylated wood. Initial brown-rot decay in acetylated wood exhibited a different trend compared to unmodified wood, with an increased OH accessibility and a significant reduction in acetyl content. This was followed by a stable, low OH accessibility and plateau in acetyl content above 10% mass loss in acetylated wood. In unmodified wood, the OH accessibility was nearly constant throughout decay, while the initially low acetyl content decreased linearly with mass loss. ATR-FTIR analysis confirmed the differences in acetyl removal between unmodified and acetylated wood. Wood-water relations before and after brown-rot decay were determined with low-field nuclear magnetic resonance (LFNMR) relaxometry on water saturated samples. For the decayed acetylated wood, the behaviour of the water corresponded well with de-acetylation observed by chemical characterization. The results show that after removal of acetyl groups, degradation of acetylated wood by R. placenta occurred at a similar rate to that of unmodified wood.

Til dokument

Sammendrag

This paper discusses the applicability of the Flory–Huggins and Vrentas sorption models for studying the sorption behaviour of wood. This theory was originally developed to explain the sorption behaviour of glassy polymers and was further extended to account for hysteresis effects. The model also has the advantage that, in principle, it does not require adjustable parameters for fitting and can be calculated independently of the sorption isotherm data. It was tested against some sorption isotherms and satisfactory fits to the data were obtained for both the absorption and desorption data. The values of the parameters required for satisfactory fitting were realistic, except for the magnitude of the glass transition temperature of water. As far as the authors are aware, this is the first reported study of the use of the Vrentas model to explain sorption and hysteresis in wood.