Statistical Models for Predicting Oil Composition from Hydrothermal Liquefaction of Biomass

Abstract

We used 352 published data points to develop multivariate linear regression, regression tree, and random forest models that predict the chemical composition of light oil from hydrothermal liquefaction of biomass. The mean absolute error calculated from ten-fold cross-validation indicates the random forest model had the best predictive ability, followed by regression tree and multivariate linear regression models. The random forest method is also more scalable than multivariate linear regression for data points outside the range of the dataset. The decision tree methods yield minimal information for improving understanding of the HTL process chemistry. Multivariate linear regression, on the other hand, identified previously unknown ternary interactions. For example, interactions involving lipid, lignin, and protein increase the abundance of N-containing compounds in the light oil. Further experimentation with lipid, lignin, and protein model compounds showed the formation of large amounts of undesirable long-chain amides in oil. This work shows that using multiple statistical models can further deepen the understanding of the HTL process in addition to providing tools that predict process outcomes.

Unravelling the Role of Biochemical Compounds within the Hydrothermal Liquefaction Process of Real Sludge Mixtures

“…This increase, which is more marked in the water fractions with respect to the bio-crude, could be correlated to higher levels of proteins (see also Table 1) and tannins typically present in TS (cf. also reaction schemes reported in [39,52]).…”

“…The da ported in Figure 5 confirm the great abundance of aliphatic compounds and reveale a higher % of TS in the sludge mixtures corresponds to the formation of greater quan of alcohols, amines, olefins and aromatic compounds compared to aliphatic compo This increase, which is more marked in the water fractions with respect to the bio-c could be correlated to higher levels of proteins (see also Table 1) and tannins typ present in TS (cf. also reaction schemes reported in [39,52]). Relative abundance of aliphatic compounds, alcohols, amines, olefins and aromati pounds in the bio-crude (A) and water (B) fractions for MS, TS and mixed sludges.…”

“…The results demonstrate that the different macro-compositions of the mixtures linked to the different contents of lipids, proteins and carbohydrates in the starting sludges significantly influence the yields of the HTL products. This is due to the complexity of the reaction patterns involved in the HTL process which are inevitably driven by the initial composition of the system [39]. Table 2 reports the L:P:C (lipid/protein/carbohydrate) ratio for the five cases examined.…”

“…Contextually, several experimental and modelling studies based on model compounds have been reported in the literature to gain deeper knowledge on the influence of biomass macro-components on the yield and quality of bio-crude (and, consequently, on the prevailing reaction pathways), also highlighting the possible synergistic or antagonistic effects [39][40][41][42]. Chacón-Parra et al [40] investigated Maillard reactions during HTL by adopting a central composite design to validate the impact of the carbohydrate/protein mass ratio on the process performance at different temperature and residence time.…”

Characterization of Biocrude Produced by Hydrothermal Liquefaction of Municipal Sewage Sludge in a 500 mL Batch Reactor

“…Therefore, the use of sewage sludge is convenient both to lower the economic impact of waste disposal in wastewater treatment plants and to increase the economic sustainability of the HTL process. The main bioconstituents of sewage sludge are proteins, carbohydrates, and lipids that under hydrothermal conditions are converted to bio-oils with high viscosity and relatively high levels of heteroatoms. − Studies on HTL of sewage sludges have been recently reviewed by Fan et al, reporting biocrude yield, HHV, and energy recovery values falling in the ranges of 10–50%, 27–75 MJ/kg, and 20–83%, respectively, as a function of the parent substrate, HTL operating conditions, and adopted catalyst . As an example, Thomsen et al applied the HTL process to sludge and wet wastes in a 20 mL batch reactor at different temperatures, also investigating the mixing of the different biomasses (fat, oil and grease, food waste, primary sludge, and secondary sludge) as feedstock .…”