Hydrothermal humification of organic waste

Abstract

This invention discloses a method for hydrothermal humification of organic waste. The organic waste is mixed with potassium hydroxide and water, and subjected to a hydrothermal reaction. The organic waste contains 25-45 wt% cellulose, 10-25 wt% hemicellulose, 10-25 wt% lignin, and 8-12 wt% protein, respectively, and has a lipid content not exceeding 5 wt%. The method provided by this invention determines the optimal feed conditions and hydrothermal reaction conditions under mesophilic rapid hydrothermal humification, improving the humic acid yield, which can reach up to 19 wt% of the feed dry weight. This invention achieves value-added recycling of biomass and effectively reduces process energy and pesticide consumption.

Description

Technical Field

[0001] This invention belongs to the field of organic waste value-added recycling technology, specifically relating to a method for hydrothermal humification of organic waste. Background Technology

[0002] Humic acid (HA) and fulvic acid (FA) are the main components of humus, possessing multiple important functions such as soil improvement, plant growth promotion, and heavy metal chelation, and are widely used in agriculture and the environment. Currently, humic acid production mainly relies on extraction from coal or traditional composting methods. However, coal is a non-renewable resource, and its extraction process poses environmental pressures; while traditional composting methods suffer from drawbacks such as long cycles, low efficiency, and unstable product quality. In contrast, hydrothermal humification technology, as an emerging method, can rapidly convert biomass-rich organic waste (such as straw, sludge, and kitchen waste) into humic acid in a high-temperature, high-pressure subcritical aquatic environment, realizing the resource utilization of waste. It has significant advantages such as fast reaction speed and clean process, demonstrating enormous application potential.

[0003] Currently, hydrothermal humification research has utilized various biomass raw materials, such as lignocellulosic biomass, municipal sludge, and livestock manure. Studies have shown that the main components of biomass, such as cellulose, hemicellulose, lignin, protein, and lipids, are key precursors for the formation of humic acid and fulvic acid. Under different hydrothermal reaction conditions (such as temperature and time) and different initial ratios, these components undergo a series of complex reactions, including hydrolysis, dehydration, and condensation, ultimately resulting in varying yields of humic acid. Publicly available research indicates that lignin, due to its aromatic structure, is generally considered a good precursor to humic acid, and the Maillard reaction between proteins and carbohydrates is also an important pathway for humic material formation. However, existing research largely focuses on the process optimization of single or mixed raw materials, generally lacking systematic and quantitative studies on the precise compatibility of core raw material components (cellulose, hemicellulose, lignin, protein, and lipids) and the impact of optimized hydrothermal humification reaction conditions on humification potential. The lack of this core compatibility relationship has led to the development of hydrothermal humification technology relying heavily on time-consuming and labor-intensive trial-and-error experiments. It is difficult to quickly and economically determine the optimal reaction conditions for specific complex raw materials, resulting in a huge waste of human and material resources. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method for hydrothermal humification of organic waste, which addresses the shortcomings of the prior art.

[0005] Another technical problem that this invention aims to solve is to provide a hydrothermal product.

[0006] The final technical problem to be solved by the present invention is to provide the application of the hydrothermal product in the preparation of soil conditioners.

[0007] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows:

[0008] The first aspect of this invention provides a method for hydrothermal humification of organic waste, wherein the organic waste is mixed with potassium hydroxide and water and subjected to a hydrothermal reaction.

[0009] The organic waste contains 25-45 wt% cellulose, 10-25 wt% hemicellulose, 10-25 wt% lignin and 8-12 wt% protein, respectively, and the lipid content is not higher than 5 wt%.

[0010] The amount of potassium hydroxide added is 6.5 to 7.0 wt% of the organic waste.

[0011] The hydrothermal reaction is carried out in a hydrothermal reactor, which is heated to 190-195°C and then held at that temperature for 85-90 minutes; the heating rate of the hydrothermal reactor is no higher than 5°C / min.

[0012] In some embodiments, the organic waste is mixed with water at a mass ratio of 1:6 to 1:10.

[0013] In some embodiments, the organic waste is dried kitchen waste three-phase separated solid residue and / or dried Chinese medicine residue; the main components of the Chinese medicine residue are astragalus, licorice and phellodendron bark, etc.

[0014] In some embodiments, the contents of cellulose, hemicellulose, lignin, protein, and lipid in the dried kitchen waste three-phase separation solid residue are 32.28 wt%, 23.04 wt%, 11.42 wt%, 13.22 wt%, and 2.17 wt%, respectively; and the contents of cellulose, hemicellulose, lignin, protein, and lipid in the dried traditional Chinese medicine residue are 25.81 wt%, 17.53 wt%, 13.52 wt%, 8.29 wt%, and 0.66 wt%, respectively.

[0015] In some embodiments, the amount of dried kitchen waste three-phase separation solid residue and / or dried Chinese medicine residue added to the organic waste is adjusted so that the contents of cellulose, hemicellulose, lignin and protein in the organic waste are 25-45 wt%, 10-25 wt%, 10-25 wt% and 8-12 wt%, respectively, and the lipid content is not higher than 5 wt%.

[0016] The second aspect of the present invention provides a hydrothermal product prepared by the method described.

[0017] The humic acid content in the hydrothermal products is 1.8 to 1.9 wt%.

[0018] A third aspect of the present invention provides the application of the aforementioned hydrothermal product in the preparation of soil conditioners.

[0019] The hydrothermal products can be used directly as soil conditioners; alternatively, the hydrothermal products can be subjected to solid-liquid separation to obtain liquid and solid products, which can be used as soil conditioners respectively; or humic acid can be extracted from the liquid and solid products to prepare soil conditioners.

[0020] The humic acid content in the liquid phase product and the solid phase product accounts for 10-11 wt% and 89-90 wt% of the total humic acid content in the hydrothermal product, respectively.

[0021] In the method for hydrothermal humification of organic waste of the present invention, the requirements for the composition of the feed are obtained by predicting the hydrothermal humification potential of biomass through machine learning, which yields the feed requirements that maximize the humic acid (HA) yield. The method includes the following steps:

[0022] S1: Data on hydrothermal humification published between 2019 and 2025 were collected from platforms such as Web of Science, Science Direct, and SCOPUS. The dataset contains n>300 sets, including 10 factor variables (cellulose content, hemicellulose content, lignin content, protein content, lipid content, reaction temperature, reaction time, feed solid-liquid ratio, additive type, and additive addition ratio) as inputs and HA yield as output. After data preprocessing, the dataset is divided into training and test sets.

[0023] S2: Five classic models were selected: Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Machine (GBM), and Ridge Regression (RR). A hydrothermal humification potential prediction model was built using Python Scikit-learn. A grid search strategy was employed to optimize hyperparameters and reduce overfitting, with a focus on optimizing R... 2 Model with a value >0.85.

[0024] S3: Based on the coefficient of determination (R²) 2The optimal model was determined using metrics such as root mean square error (RMSE) and mean absolute error (MAE), and presented in Taylor diagram form using the cosine law for model interpretation and feature importance analysis. The SHapley Additive exPlanations (SHAP) method was employed and combined with the machine learning model for model interpretation and feature importance analysis.

[0025] S4: Bayesian optimization was used to tune the hyperparameters of the optimal model, and 10-fold cross-validation was used to select the optimal parameters. An optimized XGBoost model was constructed to derive the feed requirements for the highest humic acid yield. Based on previous research, the model recommends the following precise feed requirements for low-temperature hydrothermal humification: cellulose, hemicellulose, lignin, protein, and lipid contents of 25–45 wt%, 10–25 wt%, 10–25 wt%, 8–12 wt%, and ≤ 5 wt%, respectively.

[0026] By systematically applying machine learning techniques to the prediction and optimization of the hydrothermal humification potential of biomass, high-precision prediction of humic acid (HA) yield has been achieved. Multiple machine learning models were constructed to reveal the complex nonlinear mapping relationship between the biomass core component compatibility, hydrothermal reaction conditions, and humic acid yield, thus overcoming the bottleneck of traditional experimental screening. This method can efficiently and accurately predict the humification potential of a given biomass feedstock under hydrothermal conditions and, in turn, guide the precise compatibility of feedstocks and the intelligent optimization of process parameters. Compared with experimental methods for optimizing hydrothermal humification process conditions, this method significantly reduces the blind spots of experimental exploration, shortens the R&D cycle, and reduces energy and chemical consumption. By improving the yield of the target product, it significantly promotes the industrial application of hydrothermal humification technology for organic waste and improves the level of recycling, which is conducive to achieving high-value and high-quality resource utilization of biomass.

[0027] Beneficial effects:

[0028] The method for hydrothermal humification of organic waste proposed in this invention determines the optimal feeding and hydrothermal reaction conditions for biomass under mesophilic hydrothermal humification, thereby increasing the HA yield, which can reach up to 19 wt% of the feed dry weight. This invention achieves value-added recycling of biomass and effectively reduces process energy and pesticide consumption. Attached Figure Description

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0030] Figure 1 This is the approach to predicting the hydrothermal humification potential of biomass based on machine learning, as described in Example 1.

[0031] Figure 2 This is the data distribution of biomass hydrothermal humification studies in the literature over the past 5 years, summarized in Example 1.

[0032] Figure 3 This is an evaluation of the prediction of HA yield by different machine learning models in the hydrothermal humification system in Example 1.

[0033] Figure 4 Figure 1 shows the evaluation results of different models predicting the hydrothermal humification potential of biomass in Example 1; Figure a shows the evaluation results of the training set, and Figure b shows the evaluation results of the test set.

[0034] Figure 5 This is a flowchart illustrating the biomass hydrothermal humification and extraction of humic acid and fulvic acid according to the present invention.

[0035] Figure 6 The figures show the results of single-factor and response surface design experiments on hydrothermal humification of biomass in Example 2; Figure a shows the HA yield and HA selectivity at different hydrothermal reaction temperatures; Figure b shows the HA yield and HA selectivity at different hydrothermal reaction times; Figure c shows the HA yield and HA selectivity at different KOH addition amounts; and Figure d shows the results of the response surface design experiments.

[0036] Figure 7 The results are experimental results of hydrothermal humification under different feed ratios in Example 2.

[0037] Figure 8 This is a statistical graph showing the product generation kinetics of the hydrothermal humification of biomass in Example 3.

[0038] Figure 9 This is a comparison between the predicted values ​​of the optimized XGBoost model and the actual experimental values ​​in Example 4. Detailed Implementation

[0039] The present invention will be further described below with reference to the embodiments described below. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the present invention.

[0040] Example 1

[0041] The development and training of a machine learning model for predicting the hydrothermal humification potential of biomass includes the following steps:

[0042] Step 1: Collect hydrothermal humification literature data from platforms such as Web of Science, Science Direct, and SCOPUS. The dataset contains n>300 sets, including 10 factor variables (cellulose content, hemicellulose content, lignin content, protein content, lipid content, reaction temperature, reaction time, feed solid-liquid ratio, additive type, and additive addition ratio) as inputs and HA yield as output. After data preprocessing, the dataset is divided into training and test sets.

[0043] Step 2: Select five classic models: Extreme Gradient Boosting (XGBoost, hereinafter referred to as XGB), Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Machine (GBM), and Ridge Regression (RR). Use Python Scikit-learn to build a hydrothermal humification potential prediction model. Employ a grid search strategy to optimize hyperparameters and reduce overfitting, focusing on optimizing R. 2 Model with a value >0.85.

[0044] Step 3: Based on the coefficient of determination (R²) 2 The optimal model was determined using five indicators: root mean square error (RMSE), mean absolute error (MAE), mean square error (MSE), and mean relative error (MRE). The model was then presented in Taylor chart form using the cosine law for model interpretation and feature importance analysis.

[0045] Step 4: Use Bayesian optimization to tune the hyperparameters of the optimal model, and combine 10-fold cross-validation to select the optimal parameters. Construct the optimized XGBoost model and obtain the range of the contents of each component in the feed when the HA yield is the highest.

[0046] Based on previous research results, the model recommends that the precise feeding requirements for low-temperature hydrothermal humification are: cellulose, hemicellulose, lignin, protein and lipid contents of 25 ~ 45 wt%, 10 ~ 25 wt%, 10 ~ 25 wt%, 8 ~ 12 wt% and ≤ 5 wt%, respectively.

[0047] Figure 1 This paper demonstrates the methods and ideas for predicting the hydrothermal humification potential of biomass based on machine learning, mainly divided into three parts: data collection and preprocessing, machine learning model construction, evaluation and interpretation, and experimental verification and feedback. Figure 2 As can be seen from the box plot, current research on biomass hydrothermal humification mostly focuses on reaction temperatures around 200℃, with reaction times ranging from several minutes to several hours. Research on additive types and solid-liquid ratios is relatively concentrated, primarily because overly complex additives and excessively high solid-liquid ratios cannot be used for subsequent industrial-scale production. Furthermore, alkaline additives are chosen because they are beneficial for breaking down lignin structures and dissolving HA. Figure 3 and Figure 4 It can be concluded that the R of the training set 2 The values ​​are 0.937 (XGB), 0.904 (RF), 0.903 (GBM), 0.479 (RR), and 0.451 (SVM), respectively. The R values ​​corresponding to the test set are... 2 The values ​​were 0.915 (XGB), 0.888 (RF), 0.855 (GBM), 0.586 (RR), and 0.675 (SVM), respectively. Among the five machine learning models constructed, the XGBoost model had the best prediction results for the hydrothermal humification potential of biomass, i.e., it had smaller MAE / RMSE / MSE / MRE values ​​and a larger R value. 2 Value. Taylor diagram ( Figure 3 f) shows that the XGBoost model is closest to the ideal solution. These features indicate that the model has a high fit and can accurately predict the hydrothermal humification potential of biomass. The model recommends the following precise feed requirements for low-temperature hydrothermal humification: cellulose, hemicellulose, lignin, protein, and lipid contents of 25–45 wt%, 10–25 wt%, 10–25 wt%, 8–12 wt%, and ≤ 5 wt%, respectively.

[0048] Example 2

[0049] Based on the results of Example 1, the feed ratio was calculated, and it was found that the feed ratio with the highest HA yield in hydrothermal humification using three-phase separated solid slag from kitchen waste and traditional Chinese medicine residue was 73 wt% kitchen waste three-phase separated solid slag + 27 wt% traditional Chinese medicine residue. Based on this, a series of single-factor experiments (reaction temperature, reaction time, and KOH addition ratio) and response surface methodology (three-factor, three-level) were conducted to optimize the reaction conditions favorable for HA production. The experimental design is shown in Tables 1 and 2.

[0050] (1) Single-factor optimization experiment of hydrothermal humification conditions

[0051] ① The determination of the optimal hydrothermal reaction temperature for the hydrothermal humification of biomass organic waste includes the following steps:

[0052] Step 1: The feed scheme is 73 wt% kitchen waste three-phase separation solid residue + 27 wt% traditional Chinese medicine residue. That is, 3.65 g of dried kitchen waste three-phase separation solid residue, 1.35 g of dried traditional Chinese medicine residue, 0.35 g of potassium hydroxide (KOH) and 50 mL of deionized water are sealed in a hydrothermal reactor.

[0053] Step 2: Heat the hydrothermal reactor to different temperatures (130℃, 160℃, 190℃, 220℃ and 250℃) at a heating rate of 5℃ / min, hold for 90 min and then cool naturally to room temperature, separate the solid and liquid phases, and collect the solid residue and liquid product.

[0054] Step 3: Repeatedly centrifuge and wash the solid residue with an extraction solution containing 0.1 M NaOH and 0.1 M sodium pyrophosphate (Na4P2O7) until the supernatant becomes colorless, and collect all the supernatant.

[0055] Step 4: Add 6 M HCl to the supernatant and liquid product respectively to adjust the pH to 2.0, induce humic acid precipitation, separate the solid and liquid, collect the humic acid (HA) solid, dry and weigh it, pass the liquid through a DAX-8 resin adsorption column, and then wash it with 0.1 M NaOH solution. Analyze the fulvic acid (FA) concentration using a TOC detector.

[0056] Figure 5 The flowchart shows the hydrothermal humification of biomass and the extraction of humic acid and fulvic acid.

[0057] ② The optimal hydrothermal reaction time for the hydrothermal humification of biomass organic waste was determined using the same method and steps as in Example ①, except that the hydrothermal reaction temperature was set to 190 ℃, and the effects of different hydrothermal reaction times (30, 60, 90, 120 and 180 min) on the HA yield were investigated.

[0058] ③ The method and steps for determining the optimal amount of KOH added for hydrothermal humification of biomass organic waste are the same as those in Example ①, except that the hydrothermal reaction temperature is set to 190℃, and the effects of different amounts of KOH added (0, 3 wt%, 7 wt%, 10 wt%, and 15 wt% of organic waste mass) on HA yield are investigated.

[0059] ④ The determination of the optimal feed ratio for hydrothermal humification of biomass organic waste is the same as that in Example ①, except that the hydrothermal reaction temperature is set to 190 ℃. The effects of different feed ratios (100 wt% Chinese herbal medicine residue (ratio I), 27 wt% kitchen waste three-phase separation solid residue + 73 wt% Chinese herbal medicine residue (ratio II), 50 wt% kitchen waste three-phase separation solid residue + 50 wt% Chinese herbal medicine residue (ratio III), 73 wt% kitchen waste three-phase separation solid residue + 27 wt% Chinese herbal medicine residue (ratio IV) and 100 wt% kitchen waste three-phase separation solid residue (ratio V)) on HA yield are investigated. In this experiment, under the conditions of 73 wt% kitchen waste three-phase separation solid residue + 27 wt% Chinese herbal medicine residue (ratio IV), after hydrothermal humification reaction, 47 mL of liquid product and 2.52 g of solid residue (dry weight) were collected. The dry weight of humic acid extracted from the solid residue was 0.848 g, and the dry weight of humic acid extracted from the liquid product was 0.098 g.

[0060] The experimental results are shown in Table 1 and Figure 6As shown in figures a to c, HA selectivity refers to the percentage of HA extracted from the solid residue by mass relative to the dry weight of the solid residue, representing the degree of transformation of the solid material into the target product HA. In the single-factor experiments, the HA yield showed a trend of first increasing and then decreasing with increasing reaction temperature, reaction time, and KOH addition ratio. The point where the highest HA yield occurred may be located near (190 ℃, 90 min, 7 wt%). Further three-factor, three-level response surface methodology experiments were conducted to determine the optimal reaction conditions.

[0061] Table 1. Single-factor experimental design and results

[0062]

[0063] Note: HA yield = humic acid mass / feed dry weight × 100%

[0064] (2) Three-factor, three-level response surface experiment

[0065] The experimental procedure is the same as in Example (1), except for the hydrothermal reaction temperature, hydrothermal reaction time, and the amount of KOH added. The specific experimental design and results are shown in Table 2. The response surface plot is drawn based on the data in Table 2. Figure 6 As shown in d. The three-factor, three-level response surface design results indicate that the hydrothermal reaction optimum is located at (193.6 ℃, 88.3 min, 6.9 wt%).

[0066] At the same time, the accuracy of the machine learning model's predictions under different feed ratios was also tested, and the results are as follows: Figure 7 As shown in the figure, the HA selectivity of the solid residue source refers to the percentage of HA extracted from the solid residue by mass relative to the dry weight of the solid residue, representing the degree of transformation of the solid material into the target product HA. Under the optimal feed ratio IV recommended by the model (i.e., 73% kitchen waste three-phase separation solid residue + 27% traditional Chinese medicine residue), the highest actual HA yield of 18.92% was obtained, which is a significant improvement compared to the actual HA yield of other feed ratios. In addition, it was found that the actual HA yield was significantly higher than the theoretical HA yield calculated based on the feed composition, indicating that the kitchen waste three-phase separation solid residue and traditional Chinese medicine residue produced an excellent synergistic effect at this ratio. Therefore, the aforementioned machine learning XGB model of this invention can be considered to have good accuracy.

[0067] Table 2. Three-factor, three-level response surface experimental design and results

[0068]

[0069] Example 3

[0070] The kinetics of hydrothermal humification of biomass organic waste was analyzed, including the following steps:

[0071] Step 1: Seal 3.65 g of dried kitchen waste three-phase separation solid residue, 1.35 g of dried Chinese medicine residue, 0.35 g of potassium hydroxide (KOH) and 50 mL of deionized water in a hydrothermal reactor.

[0072] Step 2: Heat the hydrothermal reactor to 190°C at a heating rate of 5°C / min.

[0073] Step 3: Select fixed time points (0, 2, 5, 10, 20, 30, 45, 60, 75 and 90 min). Immediately after the time point, open the heating jacket of the hydrothermal reactor. After the hydrothermal product cools naturally to room temperature, separate the solid and liquid phases and collect the solid residue and liquid product.

[0074] Step 4: Repeatedly centrifuge and wash the solid residue with an extraction solution containing 0.1 M NaOH and 0.1 M sodium pyrophosphate (Na4P2O7) until the supernatant becomes colorless, and collect all the supernatant.

[0075] Step 5: Add 6 M HCl to the supernatant and liquid product respectively, adjust the pH to 2.0, induce humic acid precipitation, separate the solid and liquid, collect the humic acid (HA) solid, dry and weigh it, pass the liquid through a DAX-8 resin adsorption column, and then wash it with 0.1 M NaOH solution. Analyze the fulvic acid (FA) concentration using a TOC detector.

[0076] Figure 8 The kinetics of product formation during hydrothermal humification of biomass are illustrated. In the figure, the yield of HA from the solid residue source refers to the percentage of HA extracted from the solid residue by mass relative to the dry weight of the solid residue; the yield of total HA refers to the percentage of the total mass of HA extracted from both the solid residue and liquid products relative to the dry weight of the feed; and the fulvic acid concentration refers to the concentration of fulvic acid in the liquid products. The results show that, under the same conditions, the FA concentration in the liquid products increases steadily with time, while the total HA yield increases rapidly in the first 10 minutes of the reaction, then the increase slows down. However, the yield of HA from the solid residue source increases steadily and rapidly with increasing hydrothermal reaction time. This indicates that extending the hydrothermal reaction time mainly promotes the conversion of solid residue into HA products, while having a relatively small impact on the FA concentration in the liquid products.

[0077] Example 4

[0078] Under the optimal feed ratio, hydrothermal humification experiments were conducted using single-factor combined with response surface methodology to select the reaction conditions favorable for HA production. The results of the machine learning model were verified using experimental real values, and the results were used to correct the model.

[0079] The optimization method based on the XGBoost model includes the following steps:

[0080] Step 1: Using the SHapley Additive exPlanations (SHAP) based on game theory principles, we provide local and global explanations for the "black box model," ranking the importance of each input feature and analyzing its contribution (positive / negative impact) to the output variable.

[0081] Step 2: Select the top 5 to 8 core features based on their SHAP absolute mean. Identify 1 to 2 pairs of features with significant interactions through SHAP interaction value analysis or domain experience.

[0082] Step 3: Within a reasonable framework, representative input features and operational parameters (cellulose content, hemicellulose content, lignin content, protein content, lipid content, reaction temperature, reaction time, feed solid-liquid ratio, additive type, and additive ratio) are converged to obtain practical data for validation. The newly obtained partial validation dataset (Dataset_V) is randomly merged with the original training dataset to form an expanded training set. An incremental learning strategy is used to update the XGBoost model, and the model performance is evaluated again.

[0083] Step 4: The improved performance of the new model in predicting additional new data indicates the effectiveness of the feature selection process; otherwise, hyperparameter tuning is required.

[0084] The application of the optimized machine learning model in predicting the hydrothermal humification potential of biomass organic waste includes the following steps:

[0085] Step 1: Input the hydrothermal humification reaction parameters from Examples 2 to 3, along with additional newly collected data on the hydrothermal humification reaction (feed ratio, hydrothermal reaction temperature, hydrothermal reaction time, KOH addition amount, and solid-liquid ratio), into the optimized XGBoost model and output the model prediction data.

[0086] Step 2: Compare the actual values ​​with the corresponding model predictions using SPSS software to calculate the R-squared value between the two sets of data. 2 And MSE.

[0087] A comparison of the optimized XGBoost model predictions with the experimental results is shown in the figure. Figure 9 The XGBoost model shows excellent performance on expanded training sets. 2 R (0.979) and a very small MSE (2.368) demonstrate excellent prediction performance, particularly for experimental data from single-factor and response surface design trials. Furthermore, it exhibits excellent predictive performance even on entirely new datasets. 2The MSE values ​​were 0.901 and 2.285, respectively. This demonstrates that the approach provided by this invention can greatly reduce the blind spots in experimental exploration, shorten the research and development cycle, and effectively promote the industrial application of hydrothermal humification technology for organic waste and improve the level of recycling.

[0088] This invention provides a method and approach for the hydrothermal humification of organic waste. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for hydrothermal humification of organic waste, characterized in that, The organic waste is mixed with potassium hydroxide and water to carry out a hydrothermal reaction, yielding a hydrothermal product. The organic waste contains 25-45 wt% cellulose, 10-25 wt% hemicellulose, 10-25 wt% lignin and 8-12 wt% protein, respectively, and the lipid content is not higher than 5 wt%.

2. The method according to claim 1, characterized in that, The amount of potassium hydroxide added is 6.5 to 7.0 wt% of the organic waste.

3. The method according to claim 1, characterized in that, The hydrothermal reaction is carried out in a hydrothermal reactor, which is heated to 190-195°C and then held at that temperature for 85-90 minutes; the heating rate of the hydrothermal reactor is no higher than 5°C / min.

4. The method according to claim 1, characterized in that, The organic waste is mixed with water at a mass ratio of 1:6 to 10.

5. The method according to claim 1, characterized in that, The organic waste is dry kitchen waste three-phase separated solid residue and / or dry Chinese medicine residue.

6. The method according to claim 5, characterized in that, The amount of dried kitchen waste three-phase separation solid residue and / or dried Chinese medicine residue added to the organic waste is adjusted so that the contents of cellulose, hemicellulose, lignin and protein in the organic waste are 25 ~ 45 wt%, 10 ~ 25 wt%, 10 ~ 25 wt% and 8 ~ 12 wt%, respectively, and the lipid content is not higher than 5 wt%.

7. The hydrothermal product prepared by the method according to any one of claims 1 to 6.

8. The hydrothermal product according to claim 7, characterized in that, The humic acid content in the hydrothermal products is 1.8~1.9 wt%.

9. The use of the hydrothermal product according to claim 7 in the preparation of soil conditioners.

10. The application according to claim 9, characterized in that, The hydrothermal products can be used directly as soil conditioners; alternatively, the hydrothermal products can be subjected to solid-liquid separation to obtain liquid and solid products, which can be used as soil conditioners respectively; or humic acid can be extracted from the liquid and solid products to prepare soil conditioners.

Images