Van Krevelen diagrams based on machine learning visualize feedstock-product relationships in thermal conversion processes.

Feedstock properties play a crucial role in thermal conversion processes, where understanding the influence of these properties on treatment performance is essential for optimizing both feedstock selection and the overall process. In this study, a series of van Krevelen diagrams were generated to illustrate the impact of H/C and O/C ratios of feedstock on the products obtained from six commonly used thermal conversion techniques: torrefaction, hydrothermal carbonization, hydrothermal liquefaction, hydrothermal gasification, pyrolysis, and gasification. Machine learning methods were employed, utilizing data, methods, and results from corresponding studies in this field. Furthermore, the reliability of the constructed van Krevelen diagrams was analyzed to assess their dependability. The van Krevelen diagrams developed in this work systematically provide visual representations of the relationships between feedstock and products in thermal conversion processes, thereby aiding in optimizing the selection of feedstock and the choice of thermal conversion technique.

Exposing and understanding synergistic effects in co-pyrolysis of biomass and plastic waste via machine learning.

During co-pyrolysis of biomass with plastic waste, bio-oil yields (BOY) could be either induced or reduced significantly via synergistic effects (SE). However, investigating/ interpreting the SE and BOY in multidimensional domains is complicated and limited. This work applied XGBoost machine-learning and Shapley additive explanation (SHAP) to develop interpretable/ explainable models for predicting BOY and SE from co-pyrolysis of biomass and plastic waste using 26 input features. Imbalanced training datasets were improved by synthetic minority over-sampling technique. The prediction accuracy of XGBoost models was nearly 0.90 R2 for BOY while greater than 0.85 R2 for SE. By SHAP, individual impact and interaction of input features on the XGBoost models can be achieved. Although reaction temperature and biomass-to-plastic ratio were the top two important features, overall contributions of feedstock characteristics were more than 60 % in the system of co-pyrolysis. The finding provides a better understanding of co-pyrolysis and a way of further improvements.

Conversion of tobacco processing waste to biocrude oil via hydrothermal liquefaction in a multiple batch reactor.

Fossil fuels are the primary energy source of almost all societies and economies, but it is finite and scarce. The use of non-renewable fossil fuels threatens earth's environment. At the same time, waste from agricultural and industrial activities is increasing. Most of this waste is discarded or poorly managed, causing many other environmental issues. Converting waste to energy is a promising route to address these challenges. We investigated the hydrothermal liquefaction (HTL) of high moisture content, tobacco-processing waste in a multiple batch thermal reactor to produce biocrude oil. The effects of operating conditions were studied and optimized for maximum liquid biocrude oil yield. HTL operating conditions considered were temperatures from 280 to 340 °C and residence times from 15 to 45 min for a fixed ratio of biomass to deionized water of 1:3. The reaction temperature was found to affect the yields and distribution of products significantly. The maximum yield of the liquid biocrude oil obtained was more than 52% w/w at 310 °C and 15 min. Under these conditions, almost 90% of the energy was recovered in biocrude oil and solid products. The liquid fraction was mainly composed of phenols, ketones, and nitrogenous compounds. This study provides a potential framework for eco-technologies for biomass waste-to-energy conversion with respect to converting tobacco processing residues to liquid biofuels and biochemicals.

Machine learning prediction of biocrude yields and higher heating values from hydrothermal liquefaction of wet biomass and wastes.

Machine learning (ML) approach was applied for the prediction of biocrude yields (BY) and higher heating values (HHV) from hydrothermal liquefaction (HTL) of wet biomass and wastes using 17 input features from feedstock characteristics (biological and elemental properties) and operating conditions. Several novel ML algorithms were evaluated, based on 10-fold cross-validation, with 3 different sets of input features. An extreme gradient boosting (XGB) model proved to give the best prediction accuracy at nearly 0.9 R2 with normal root mean square error (NRMSE) of 0.16 for BY and about 0.87 R2 with NRMSE of about 0.04 for HHV. Temperature was found to be the most influential feature on the predictions for both BY and HHV. Meanwhile, feedstock characteristics contributed to the XGB model for more than 55%. Individual effects and interactions of most important features on the predictions were also exposed, leading to better understanding of the HTL system.

Compositional analysis of bio-oils from hydrothermal liquefaction of tobacco residues using two-dimensional gas chromatography and time-of-flight mass spectrometry.

Sustainable energy from biomass is one of the most promising alternative energy sources and is expected to partially replace fossil fuels. Tobacco industries have normally rid their processing residues by landfilling or incineration, affecting the environment negatively. These residues can be used to either extract high-value chemicals or generate bio-energy via hydrothermal liquefaction. The main liquid product or bio-oil consists of highly complicated chemicals. In this work, the bio-oil from hydrothermal liquefaction of tobacco processing residues was generated in a batch reactor at biomass-to-deionized water ratio of 1:3, temperature of 310°C, and 15 min residence time, yielding the maximum liquid products for more than 50% w/w. The liquid products were analyzed, using two-dimensional gas chromatography and time-of-flight mass spectrometry (GC × GC/TOF MS). This technique allowed for a highly efficient detection of numerous compounds. From the results, it was found that hydrothermal liquefaction can cleave biopolymers (cellulose, hemicellulose, and lignin) in tobacco residues successfully. The hydrothermal liquefaction liquid products can be separated into heavy organic, light organic, and aqueous phase fractions. By GC × GC/TOF MS, the biopolymers disintegrated into low molecular weight compounds and classified by their chemical derivatives and functional groups could be detected. The major chemical derivative/functional groups found were cyclic ketones and phenols for heavy organic and light organic, and carboxylic acids and N-containing compounds for the aqueous phase. Additionally, by the major compounds found in this work, simple pathway reactions occurring in the hydrothermal liquefaction reaction were proposed, leading to a better understanding of the hydrothermal liquefaction process for tobacco residues.

Biocrude oil production via hydrothermal liquefaction of food waste in a simplified high-throughput reactor.

In this study, the hydrothermal liquefaction of food waste was investigated for generating bio-oils or biocrudes using a simplified high-throughput reactor. Different operating conditions were parametrically tested for obtaining maximum yield of biocrudes at temperatures from 280 to 340 °C and sample-to-water mixture ratios from 1:3 to 1:7, with 30 min residence time. The biocrude yields were found to increase significantly with increasing temperature and mixture ratio. A maximum biocrude yield of approximately 40% w/w dry basis (from a total liquid product yield of over 56% w/w) and energy recovery of over 70% were obtained at 340 °C and a mixture ratio of 1:7. High heating values of the resulting biocrude and solid residue were remarkably better than those of the raw material. Fatty acids, amides, N-containing compounds, and cyclic ketones were identified to be the main components of the biocrudes using advanced GCxGC/TOF-MS.

Production and characterization of bio-oils from fast pyrolysis of tobacco processing wastes in an ablative reactor under vacuum.

Application of advanced pyrolysis processes to agricultural waste for liquid production is gaining great attention, especially when it is applied to an economic crop like tobacco. In this work, tobacco residues were pyrolyzed in an ablative reactor under vacuum. The maximum bio-oil yield of 55% w/w was obtained at 600°C with a particle size of 10 mm at a blade rotation speed of 10 rpm. The physical properties of the products showed that the oil produced was of high quality with high carbon, hydrogen, and calorific value. Two-dimensional gas chromatography/time-of-flight mass spectrometric analysis results indicated that the oils were complex mixtures of alkanes, benzene derivative groups, and nitrogen-containing compounds. In addition, 13C NMR results confirmed that long aliphatic chain alkanes were evident. The alkanes were likely converted from furans that were decomposed from hemicelluloses. Ablative pyrolysis under vacuum proved to be a promising option for generating useful amount of bio-oils from tobacco residues.

Machine learning prediction of cellulose-rich materials from biomass pretreatment with ionic liquid solvents.

Ionic liquid solvents (ILSs) have been effectively utilized in biomass pretreatment to produce cellulose-rich materials (CRMs). Predicting CRM properties and evaluating multi-dimensional relationships in this system are necessary but complicated. In this work, machine learning algorithms were applied to predict CRM properties in terms of cellulose enrichment factor (CEF) and solid recovery (SR), using 23-feature datasets from biomass characteristics, operating conditions, ILSs identities, and catalyst. Random forest algorithm was found to have the highest prediction accuracy with RMSE and R2 of 0.22 and 0.94 for CEF, as well as 0.07 and 0.84 for SR, respectively. Highly influential features on making predictions were mainly from biomass characteristics andILS treatment'soperating conditions, totally contributed 80% on CEF and 60% on SR. One- and two-way partial dependence plots were used to explain/interpret the multi-dimensional relationships of the most important features. Our findings could be applied in designing new ILSs and optimizing the process conditions.

Bio-oils from vacuum ablative pyrolysis of torrefied tobacco residues.

Fast pyrolysis, in combination with torrefaction pretreatment, was used to convert tobacco residues to value-added bio-fuels and chemicals. Tobacco plant residues were torrefied at 220, 260, and 300 °C, before being pyrolyzed at 450, 500, 550, and 600 °C in a rotating blade ablative reactor under vacuum conditions to test the effects on product yields. With torrefaction, tobacco residues thermally decomposed 20-25% w/w at low temperatures. Torrefaction and pyrolysis temperatures were found to markedly affect pyrolytic product yields of bio-chars and bio-oils, while having no effect on gas-phase products. Bio-oil yields exhibited a direct relation with pyrolysis temperature and an inverse relation with torrefaction temperature. Bio-oils produced were separated into light and heavy oils and analyzed by GC-MS, and 1H and 13C NMR. Nicotine was found to be the main compound in the light and heavy oils along with several phenols and cresols in the heavy oil.

Torrefaction of pelletized corn residues with wet flue gas.

Corn residue pellets were torrefied with wet flue gas, simulated by steam (0-21% v/v), CO2 (12% v/v), and O2 (4% v/v), balanced with N2 as reactive gas, in a fixed bed reactor at 260 °C of temperature and at 10-40 min of residence time. The distribution and yields of torrefied pellets, liquid, and gas products were examined. The microstructural changes of torrefied pellets were evaluated by Raman spectroscopy and scanning electron microscopy, while the components of gas products were analyzed by mass spectrometry. Residence time and steam concentration in the reactive gas were found to have significant effects on the products yield distribution, the porosity of the torrefied pellets, and the concentrations of CO, CH4, H2, and CO2 in the gas products. At high steam concentrations, the decomposition reaction of hemicellulose and lignin in the raw pellets, and the formation of the graphene structures in torrefied pellets occurred faster.