Automated machine learning-aided prediction and interpretation of gaseous by-products from the hydrothermal liquefaction of biomass.

Hydrothermal liquefaction (HTL) is a thermochemical conversion technology that produces bio-oil from wet biomass without drying. However, by-product gases will inevitably be produced, and their formation is unclear. Therefore, an automated machine learning (AutoML) approach, automatically training without human intervention, was used to aid in predicting gaseous production and interpreting the formation mechanisms of four gases (CO2, CH4, CO, and H2). Specifically, four accurate optimal single-target models based on AutoML were developed with elemental compositions and HTL conditions as inputs for four gases. Herein, the gradient boosting machine (GBM) performed excellently with train R2 ≥ 0.99 and test R2 ≥ 0.80. Then, the screened GBM algorithm-based ML multi-target models (maximum average test R2 = 0.89 and RMSE = 0.39) were built to predict four gases simultaneously. Results indicated that biomass carbon, solid content, pressure, and biomass hydrogen were the top four factors for gas production from HTL of biomass. This study proposed an AutoML-aided prediction and interpretation framework, which could provide new insight for rapid prediction and revelation of gaseous compositions from the HTL process.

Distribution characteristics and migration pathways of metals during hydrothermal liquefaction of municipal sewage sludge in the presence of various catalysts.

A series of hydrothermal liquefaction (HTL) experiments with two different samples of municipal sewage sludge (MSS) were conducted at 350 °C for 30 min residence time in a high pressure batch reactor. The main aim of the study was to explore the distribution and migration pathways of a broad range of metals and metalloids in the HTL products (bio-oil, char and aqueous phase) obtained in the presence of various homogeneous and heterogeneous catalysts (Na2CO3, Li2CO3, K2CO3, Ba(OH)2, Fe2O3, CeO2, NiMo/MoO3, MoS2, Ni/NiO, SnO2, FeS). The elements under study included 16 environmentally significant metals and metalloids (As, B, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, Se, Sn, Zn and Hg). The study showed that the quantitative migration of the tested metals and metalloids to the particular HTL products, relative to their initial content in the raw sludge, is different for the individual elements. Most metals exhibited a particularly strong affinity to the solid fraction (biochar). In the obtained HTL bio-oils, all tested elements were identified, except of Cd. It was also found that B and As have high affinity to the aqueous phase. A direct effect of catalysts on the contents of some elements in the products was also proved by the study, e.g. increased concentration of Cr in the biochar when Fe2O3 was used as a process catalyst. Due to the wide scope of the tested elements and broad range of catalyst used, the results obtained represent a unique and comprehensive set of environmental data compared to similar HTL studies previously conducted for MSS.

Coupling influences of organic components and temperature on nitrogen transformation and hydrochar characterization during hydrothermal carbonization of sewage sludge.

Nitrogen (N) in sewage sludge (SS) should be reduced if it is to be used to produce clean solid fuels. However, the N transformation during hydrothermal carbonization (HTC) of SS is not yet fully understood. Since the composition of SS is complex, it is wise to study a model compound, which should have typical functional groups of organic components. Hence, in this study, six model components (protein, lipid, cellulose, hemicellulose, humic acid, and lignin) representing the main organic components in SS were mixed with SS and treated at 150-270 °C for 1 h. The influence of the organic component and reaction temperature on hydrochar yield, hydrochar characterization, and N distribution in the products was investigated. Except for proteins and lipids, all the other components were found to contribute to the N content and aromatization of the hydrochar. Humus shows the best comprehensive performance in terms of both reducing the N content and increasing the aromaticity. The strongest effects of hemicellulose and cellulose on N retention in hydrochar are found to occur at 210 °C and 240 °C, respectively. The N retention caused by lignin is correlated with the Mannich reaction at 240 °C, while humus significantly promotes N transformation at 240 °C. For carbohydrates, lignin, and humus, the temperatures required for increasing the N content and aromaticity maintain a high degree of consistency. Although protein pulls down the energy recovery (ER) and yield of the hydrochar, observations indicate that it favors the carbonization process. This finding can be used for estimating the N content and quality of hydrochar and provides references for future research targeting the upgrading of hydrochar.

Characterisation of inorganic constitutions of condensate and solid residue generated from small-scale ex situ experiments in the context of underground coal gasification.

This paper deals with the characterisation of inorganic constitutions generated at various operating conditions in the context of underground coal gasification (UCG). The ex situ small-scale experiments were conducted with coal specimens of different rank, from the South Wales Coalfield, Wales, UK, and Upper Silesian Coal Basin, Poland. The experiments were conducted at various gaseous oxidant ratios (water: oxygen = 1:1 and 2:1), pressures (20 bar and 36 bar) and temperatures (650°C, 750°C and 850°C). Increasing the amount of water in the oxidants proportionately decreased the cationic elements but increased the concentrations of anionic species. The temperature played minor impact, while the high-pressure experiments at temperature optimum to produce methane-rich syngas (750°C) showed significant reduction in cationic element generation. However, both coal specimens produced high amount of anionic species (F, Cl, SO4 and NO3). The "Hard" bituminous coal from Poland produced less gasification residues and condensates than the South Wales anthracitic coal due to its higher reactivity. The inorganic composition found in the solid residue was used in the theoretical calculation to predict the dissolved product concentrations when the solid residue interacts with deep coal seam water in the event of UCG cavity flooding. It was evident from the solubility products of the Cr, Ni and Zn that changes in the groundwater geochemistry occur; hence, their transportation in the subsurface must be studied further.

In situ sorption phenomena can mitigate potential negative environmental effects of underground coal gasification (UCG) - an experimental study of phenol removal on UCG-derived residues in the aspect of contaminant retardation.

The aim of the study was to investigate the sorption interactions between phenol and materials obtained from four different underground coal gasification (UCG) ex-situ simulations. These interactions are significant in terms of the impact of the UCG on the groundwater environment. Sorption parameters were determined for two sample types: raw coal mined from the coal-bed and then subjected to the gasification process; and char residue acquired from the cavity formed as a result of the UCG processes. Laboratory-scale tests were carried out using deionized water and aqueous solutions with increasing concentrations of phenol (from 50 mg/dm3 to 2000 mg/dm3) at 298 K. On the assumption of physical interactions (non-specific physisorption) and due to a nonlinear mass distribution of adsorbed substances as a function of equilibrium concentration, the Freundlich isotherm model was applied to describe adsorption phenomena. The isotherms have good fitting (R2 from 0.5716 to 0.9811). Relatively high percentage phenol removal efficiency was observed for all tested chars (from 17.0% to 99.8% for the 1.0-2.5 mm fraction and from 6.9% to 99.6% for the 10.0-12.5 mm fraction). Additionally, the sorption characteristics was used to evaluate the retardation coefficients. The largest delay in the organic pollutant migration in the environment around a UCG reactor occurs for phenol transport in the layer of the post-process char from 'Wesola' after 40 bar pressure experiment.

Effect of ultrasound pretreatment of municipal sewage sludge on characteristics of bio-oil from hydrothermal liquefaction process.

Hydrothermal liquefaction experiments with a moist (water content 80.64 wt%) municipal sewage sludge (SS) were conducted at 280, 300, 320, 340 and 360 °C for raw and ultrasound pretreated (energy input 4500 kJ/kg of solid matter) sludge. The main aim of the experiments was to explore the effects of ultrasounds on yields and composition of bio-oil and liquefaction by-products (gas and char). The bio-oils obtained were viscous, dark brown liquids with a specific odor and heating values in the range 31.39-37.35 MJ/kg. A temperature of 320 °C was found to be the optimum for the maximum yields of the SS-derived bio-oil. During the tests a significant effect of ultrasounds on bio-oil yields was proven. A positive effect of ultrasound energy input was observed at each of the liquefaction temperatures and the maximum effect (increase of 19 wt% in bio-oil yield) was achieved at 320 °C. The GC-MS analysis revealed that the bio-oils were complex mixtures of many chemical compounds belonging to different chemical classes: aliphatic hydrocarbons, fatty alcohols, N-aliphatics, phenols, aromatic esters, aldehydes, ketones, alcohols and carboxylic acids. It was found that influence of ultrasounds on the chemical composition of the bio-oils was negligible.

Chemical and toxicological evaluation of underground coal gasification (UCG) effluents. The coal rank effect.

The effect of coal rank on the composition and toxicity of water effluents resulting from two underground coal gasification experiments with distinct coal samples (lignite and hard coal) was investigated. A broad range of organic and inorganic parameters was determined in the sampled condensates. The physicochemical tests were supplemented by toxicity bioassays based on the luminescent bacteria Vibrio fischeri as the test organism. The principal component analysis and Pearson correlation analysis were adopted to assist in the interpretation of the raw experimental data, and the multiple regression statistical method was subsequently employed to enable predictions of the toxicity based on the values of the selected parameters. Significant differences in the qualitative and quantitative description of the contamination profiles were identified for both types of coal under study. Independent of the coal rank, the most characteristic organic components of the studied condensates were phenols, naphthalene and benzene. In the inorganic array, ammonia, sulphates and selected heavy metals and metalloids were identified as the dominant constituents. Except for benzene with its alkyl homologues (BTEX), selected polycyclic aromatic hydrocarbons (PAHs), zinc and selenium, the values of the remaining parameters were considerably greater for the hard coal condensates. The studies revealed that all of the tested UCG condensates were extremely toxic to V. fischeri; however, the average toxicity level for the hard coal condensates was approximately 56% higher than that obtained for the lignite. The statistical analysis provided results supporting that the toxicity of the condensates was most positively correlated with the concentrations of free ammonia, phenols and certain heavy metals.

Analysis of the organic contaminants in the condensate produced in the in situ underground coal gasification process.

Addressing the environmental risks related to contamination of groundwater with the phenolics, benzene, toluene, ethyl benzene, xylene (BTEX) and polycyclic aromatic hydrocarbons (PAHs), which might be potentially released from the underground coal gasification (UCG) under adverse hydrogeological and/or operational conditions, is crucial in terms of wider implementation of the process. The aim of this study was to determine the main organic pollutants present in the process condensate generated during the UCG trial performed on hard coal seam in the Experimental Mine 'Barbara', Poland; 8,933 L of condensate was produced in 813 h of experiment duration (including 456 h of the post-process stage) with average phenolics, BTEX and PAH concentrations of 576,000, 42.3 and 1,400.5 µg/L, respectively. The Hierarchical Clustering Analysis was used to explore the differences and similarities between the samples. The sample collected during the first 48 h of the process duration was characterized by the lowest phenanthrene, anthracene, fluoranthene and pyrene contents, high xylene content and the highest concentrations of phenolics, benzene, toluene and ethyl benzene. The samples collected during the stable operation of the UCG process were characterized by higher concentrations of naphthalene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo(a)anthracene, chrysene, while in the samples acquired in the post-process stage the lowest concentrations of benzene, toluene, naphthalene, acenaphthene and fluorene were observed.

Chemometric Study of the Ex Situ Underground Coal Gasification Wastewater Experimental Data.

The main goal of the study was the analysis of the parameters of wastewater generated during the ex situ underground coal gasification (UCG) experiments on lignite from Belchatow, and hard coal from Ziemowit and Bobrek coal mines, simulated in the ex situ reactor. The UCG wastewater may pose a potential threat to the groundwater since it contains high concentrations of inorganic (i.e., ammonia nitrogen, nitrites, chlorides, free and bound cyanides, sulfates and trace elements: As, B, Cr, Zn, Al, Cd, Co, Mn, Cu, Mo, Ni, Pb, Hg, Se, Ti, Fe) and organic (i.e., phenolics, benzene and their alkyl derivatives, and polycyclic aromatic hydrocarbons) contaminants. The principal component analysis and hierarchical clustering analysis enabled to effectively explore the similarities and dissimilarities between the samples generated in lignite and hard coal oxygen gasification process in terms of the amounts and concentrations of particular components. The total amount of wastewater produced in lignite gasification process was higher than the amount generated in hard coal gasification experiments. The lignite gasification wastewater was also characterized by the highest contents of acenaphthene, phenanthrene, anthracene, fluoranthene, and pyrene, whereas hard coal gasification wastewater was characterized by relatively higher concentrations of nitrites, As, Cr, Cu, benzene, toluene, xylene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, and benzo(a)pyrene.