A method for differentiated management of pollution prevention and control in chemical park sites

Abstract

The present invention relates to differentiated management of site pollution prevention and control, and specifically to a method for differentiated management of site pollution prevention and control in a chemical park, including: determining the influencing factors of differentiated management of site pollution prevention and control in a chemical park, and evaluating the contribution of each influencing factor to the pollution of the chemical park site; identifying the environmental characteristic index factors of differentiated management of site pollution prevention and control in the chemical park by dimensionality reduction; constructing a scoring system based on the quantitative results of the factors in the chemical park site and in combination with the contribution rate, dividing the levels, and assigning points to each level; based on the level scoring results, combined with the investigation risks of the chemical park site, etc., constructing a technical route for regional zoning of pollution prevention and control, and performing differentiated management. The present invention also relates to a differentiated site grading and phased management method for site pollution prevention and control in an enterprise, which gradually judges the pollution risk, migration risk, potential risk and surrounding risk of the enterprise site to determine the site grading of the enterprise site.

Description

Technical Field

[0001] The present invention relates to differentiated management of site pollution prevention and control, and in particular to a zoning and classification technical method for differentiated management of site pollution prevention and control in a chemical park. Background Art

[0002] At present, pollution prevention and control management in chemical parks is mainly focused on the prevention and control of VOCs in the park's atmosphere. In terms of site pollution, methods are mostly established for single pollution types and site pollution transmission media. No global zoning and classification management technical solutions have been proposed for the large area and complex hydrogeological conditions of chemical parks. As a result, pollution prevention and control solutions for single sites are of limited applicability to large parks.

[0003] Therefore, it is necessary to develop a technical method for pollution prevention and control zoning and classification management with chemical parks and enterprise plots as the main bodies, so as to provide a scientific basis for the differentiated management of chemical parks and enterprise plots. Summary of the invention

[0004] Purpose of the invention: The technical problem to be solved by the present invention is to address the deficiencies in the prior art and to provide a zoning and classification technical method for differentiated management of pollution prevention and control in chemical parks and a differentiated site grading and phased management method for pollution prevention and control in enterprise sites.

[0005] Invention idea: The present invention takes into account the site pollution characteristics, hydrogeological conditions, production technology, etc., and combines environmental management needs, economic level, changing trends of surrounding sensitive receptors and subsequent planning and utilization changes, etc., to provide a scientific basis for the differentiated management of chemical parks and enterprise plots, thereby developing a pollution prevention and control zoning, classification, grading and phased management method with chemical parks and enterprise plots as the main body.

[0006] In order to solve the above technical problems, the present invention discloses the following technical solutions:

[0007] In a first aspect, the present invention discloses a zoning and classification technology method for differentiated management of pollution prevention and control at chemical park sites.

[0008] Wherein, the technical method comprises the following steps:

[0009] (1) Collect relevant information on chemical park sites, establish a basic database for the sites, clean and pre-process the data, and divide it into a modeling set and a validation set. Combined with the influence of eight major environmental factors on site pollution, including hydrogeology, soil properties, meteorological conditions, industry categories, years of production, socio-economics, regulatory levels, and pollutant properties, the influencing factors of differentiated management of chemical park site pollution prevention and control are determined through model training and test analysis, and the contribution of each influencing factor to chemical park site pollution is quantitatively evaluated;

[0010] (2) From the identified influencing factors of differentiated management of pollution prevention and control at chemical industrial parks, reduce the dimension and identify the environmental characteristic indicator factors of differentiated management of pollution prevention and control at chemical industrial parks;

[0011] (3) Based on the quantitative results of environmental characteristic index factors at the chemical park site and combined with the contribution rate, a scoring system is constructed to divide the level of each environmental characteristic index factor and assign points to each level;

[0012] (4) Based on the graded scoring results of environmental characteristic index factors, combined with any one or more of the survey risk, regional risk, historical risk and surrounding risk of the chemical park site, a technical route for regional zoning of pollution prevention and control of chemical park sites is constructed to carry out differentiated management.

[0013] In step (1), the chemical park site-related information includes social economy, chemical industry composition, park geographical location, site pollution characteristics, meteorological conditions, etc. The data cleaning includes processing missing values, outliers and duplicate values. Missing values ​​may affect the performance of the model, so they can be processed by filling, deleting or interpolating; outliers may interfere with the training of the model, so they need to be identified and properly processed. The preprocessing includes representing qualitative data types with numerical types that can be processed by the model. Statistical methods can be used to scale features, mainly including standardization (scaling features to a mean of 0 and a standard deviation of 1) and normalization (scaling features to a specific range, such as between 0 and 1). According to domain knowledge or model requirements, operations such as feature derivation, feature combination or feature conversion can be performed to extract more information and improve model performance. The modeling set and the validation set are divided by using the Kennard-Stone (KS) algorithm in a ratio of 7:3. The KS algorithm selects representative samples in the feature space based on Euclidean distance, selects samples with large feature differences into the modeling set, and the remaining samples are classified into the validation set.

[0014] In step (1), the eight categories of environmental factors include 23 environmental factors, specifically: hydrogeology including soil medium type of vadose zone, runoff potential, groundwater depth, terrain moisture index, rainfall, elevation, terrain slope; soil properties including soil bulk density, soil temperature, soil thickness; meteorological conditions including PM2.5 concentration, PM10 concentration, average relative humidity of air; industry category; production years; socio-economic including surrounding population density and production and operation area; regulatory level including soil / groundwater monitoring, air / surface water monitoring, early warning management, emergency management; pollutant properties including pollutant mobility and pollutant volatility.

[0015] In step (1), the model training and test analysis method is to integrate Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (Xgboost) and Random Forest (RF) for model training and test analysis; considering the complexity and diversity of chemical park data sets, the integrated GBDT, XGBoost and RF models are used for model training and test analysis to identify the key elements of regional chemical park characteristics, and quantitatively evaluate the contribution of each environmental characteristic to the pollution characteristics and environmental quality of the chemical park. GBDT, XGBoost and RF models are good at processing structured, high-dimensional, large-scale and complex nonlinear relationship data sets, and can effectively capture the complex relationship between the key elements of chemical parks. By learning the complex relationship between various environmental characteristics and evaluating the importance of each feature indicator when constructing a decision tree, it helps to identify the key elements of chemical park pollution characteristics.

[0016] In some embodiments, the number of iterations in the GBDT model is set to 100, the learning rate is set to 0.2, the minimum number of samples is set to 5, and other parameters are retained as default values; in the XGBoost model, the learning_rate is set to 0.01, the gamma is set to 0.1, and other parameters are retained as default values; in the RF model, the number of trees is set to 1000, the minimum number of samples of the terminal nodes is 5, and other parameters are retained as default values. During the training process of the model, the environmental feature gain may be negative, which indicates that its contribution to the analysis and identification of the characteristics of the chemical park is small and negligible. In this case, it is recommended to directly delete such environmental features and retrain the model on the screened data set. By sorting the contribution of each environmental variable in the model output results, the larger the value, the more important the variable is. Except for environmental variables with significantly larger impacts, the actual impact of the variables ranked lower in a certain element may be small or there is no direct impact, but due to some direct or indirect correlation with other variables, this variable has a certain degree of contribution rate in the importance analysis of the element. Finally, by calculating the model determination coefficient (R 2 ), prediction root mean square error (RMSE), the comprehensive evaluation of the two is more objective for the evaluation of non-normal distribution attribute prediction model. 2 The closer the RMSE is to 1, the more reliable the output of the model is.

[0017] In step (2), the dimensionality reduction identification method is principal component analysis (PCA); the environmental characteristic index factors are selected by screening the influencing factors with a cumulative contribution rate of > 85%. PCA can convert multidimensional data into a set of related variables, while helping to identify the most representative features in the data set on the basis of reducing the data dimension and reducing the computational complexity of model training. PCA reduces the dimension of features while ensuring that key factors are covered as much as possible.

[0018] In step (3), the quantitative means is to evaluate the key factors of site pollution risk and their importance based on the output results of the model, quantify the environmental characteristic indicators in the form of a percentage system; and determine the grade classification standards based on the specific characteristics of each environmental factor and its correlation with the degree of pollution. Specifically, the grade of each environmental characteristic indicator factor is divided according to the pollution rate of the environmental characteristic indicator factor in the chemical park site; and the contribution rate of the environmental characteristic indicator factor to the pollution of the chemical park site is assigned points.

[0019] In step (3), the scoring system includes three levels: environmental factor categories, environmental factor subcategories and environmental factor indicator layers. The environmental factor subcategories are the environmental characteristic indicator factors; environmental factor subcategories are subcategories of environmental factor categories, and environmental factor indicator layers are subcategories of environmental factor categories; environmental factor categories may include production years, industry categories, hydrogeology, pollutant properties, regulatory levels, etc. Environmental factor subcategories may include vadose zone soil medium types, runoff potential, terrain moisture index, groundwater depth, rainfall, pollutant mobility, pollutant volatility, soil / groundwater monitoring, air / surface water monitoring, early warning management, etc. Environmental factor indicator layers may include fill or thin layer, gravel / sand, silt / loam, clay, etc. After the system is established, scoring is performed. First, the environmental factor subcategories are measured and scored according to the contribution rate, and the scores of the environmental factor categories are added up; then, the scores of the environmental factor indicator layers are calculated by dividing the scores specified in the environmental factor subcategories into equal parts according to the number of levels.

[0020] In step (4), the regional zoning and classification management belongs to the park level. The regional zoning is based on the level of pollution risk and the interaction between the park and sensitive ecological targets, and the different areas in the park are divided. This zoning method is mainly based on the pollution investigation risk, regional risk, historical risk, and surrounding risk of each area in the park. Specifically, the regional zoning adopts a step-by-step judgment method, including two forms: conditional judgment and comprehensive scoring; the conditional judgment is to judge any one or more of the investigation risk, regional risk, historical risk, and surrounding risk of the chemical park site; the comprehensive scoring is the scoring system described in step (3).

[0021] Among them, when the survey results exceed the control value or exceed the human health risk value, the area of ​​the chemical park is a heavily polluted area; when the total score calculated according to the scoring system is greater than 70 points, the area of ​​the chemical park is a high-risk area; when the survey quality of the area is poor or has not been surveyed (i.e., the pollution situation is unknown), and the total score calculated according to the scoring system is 50<≤70, or the production years of the chemical park site is greater than 30 years, or an environmental accident has occurred, the area of ​​the chemical park is a medium-risk area; when the total score calculated according to the scoring system is ≤50, and there are no key river basins, major drinking water sources, or ecological protection areas within 1 km of the surrounding area, the area of ​​the chemical park is a low-risk area. Further, when the total score calculated according to the scoring system is ≤50, and there are key river basins, major drinking water sources, or ecological protection areas within 1 km of the surrounding area, the area of ​​the chemical park is an ecologically sensitive area. The parameters such as 50 and 70 here are recommended values ​​based on the actual scores in the collected samples and comprehensive consideration of environmental management needs.

[0022] Among them, classified management is carried out based on the results of regional zoning and combined with the utilization scenarios of different areas in the chemical park, that is, the classified management is divided according to the regional scenarios, and comprehensive consideration is given to the management of different zones of the chemical park and environmental goals, and targeted pollution prevention and control requirements for different zones are proposed; preferably, the heavily polluted areas that cannot be repaired belong to the classified management of the storage control type; the heavily polluted areas that are retired and redeveloped belong to the classified management of the restoration and utilization type; the heavily polluted areas that are retired and idle and under production control, and the high-risk areas with pollution belong to the classified management of the pollution control type; the high-risk areas without pollution, the low-risk areas that have been developed and utilized, and the ecologically sensitive areas that have been developed and utilized belong to the classified management of the pollution prevention type; the medium-risk areas, the low-risk areas that have not been developed and utilized, and the ecologically sensitive areas that have not been developed and utilized belong to the classified management of the pollution identification type.

[0023] In a second aspect, the present invention provides a differentiated site grading and phased management method for enterprise site pollution prevention and control.

[0024] Among them, the site classification and phased implementation belongs to the enterprise level. The site classification adopts a step-by-step judgment method and is classified according to the risk information that can be judged on the land. Since the site scale classification and phased implementation are mainly aimed at the specific management requirements of each operating enterprise or legacy site, one of its important goals is to urge all responsible parties to implement systematic site pollution assessment and prevention. Therefore, the site classification indicators are more specific and detailed than the indicators of the park zoning, so that the specific responsible parties of the site can gradually improve and optimize the soil and groundwater environmental quality of the enterprise and the site on time. The phased implementation policy classifies the scenarios according to the degree of urgency of governance, comprehensively considers the different hierarchical management and environmental goals in the chemical site, and puts forward targeted pollution prevention and control requirements for different levels and phases.

[0025] Specifically, the site classification adopts a step-by-step judgment method to judge and evaluate the four key indicators of the land's pollution risk, migration risk, potential risk and surrounding risk; preferably, when the investigation result exceeds the second-category land use control value or exceeds the human health risk value or there is a serious odor, the enterprise is at the health prevention and control level; when the pollution exceeds the screening value and the pollution factor mobility M>0.01, or there are coarse sand, pebbles, fine sand, sub-sand, silt, etc. in the natural soil, the enterprise is at the migration prevention and control level; when the pollution exceeds the screening value, the land risk conceptual model is missing, the production period is>30 years, or a major environmental pollution accident has occurred within 1 kilometer of the surrounding area, the enterprise is at the investigation and identification level; when there are key river basins, ecological protection areas, drinking water sources, etc. within 500 meters around the land, the enterprise is at the protection monitoring level; otherwise it is at a low risk level.

[0026] Furthermore, based on the results of site classification and the urgency of site management of the enterprise, phased measures are implemented; preferably, management needs to be initiated for the health prevention and control level that has caused health damage, and control needs to be initiated for the health prevention and control level that has not yet caused health damage. The migration prevention and control level where the pollution risk has migrated outside the site needs long-term prevention and control, the migration prevention and control level and the investigation and identification level where there is a migration trend of pollution risk need to be monitored in real time, the investigation and identification level where the migration characteristics of pollution risk are unclear needs additional investigation, the investigation and identification level where the initial adjustment exceeds the standard needs to initiate detailed adjustment, the protection monitoring level that has not been investigated needs to initiate initial adjustment, and the low-risk level with no signs of pollution needs dynamic monitoring.

[0027] Beneficial effects:

[0028] The present invention targets the characteristics of different industries and regional environmental conditions, combines the actual needs of chemical parks, takes the zoning and classification method as a guiding principle, promotes the implementation of scientific and effective prevention and control technologies and systematic and clear management strategies, and has important practical significance for achieving the coordinated unity of regional environmental protection and economic development.

[0029] The present invention provides a zoning and classification technical method for differentiated management of site pollution prevention and control in a chemical park and a differentiated site grading and phased management method for site pollution prevention and control in an enterprise. It targets the characteristics of different industries and regional environmental conditions, combines the actual needs of the chemical park, takes the zoning, classification, grading and phased method as a guiding principle, promotes the implementation of scientific and effective prevention and control technologies and systematic and clear management strategies, reduces the subjectivity of management and improves accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0031] Figure 1 To identify and assess the causes of site contamination based on integrated machine learning model screening.

[0032] Figure 2 The key elements of site pollution are extracted by dimensionality reduction based on principal component analysis method.

[0033] Figure 3 It is the key factor characteristic of site pollution environment.

[0034] Figure 4 It is the accumulation characteristics of pollutants at different depths on the site.

[0035] Figure 5 It is the volatility correlation coefficient characteristic of pollutant migration.

[0036] Figure 6 Provide technical routes for regional zoning and classification management.

[0037] Figure 7 Provide technical routes for site classification and phased implementation. DETAILED DESCRIPTION

[0038] The present invention can be better understood according to the following examples. However, it is easy for those skilled in the art to understand that the contents described in the examples are only used to illustrate the present invention, and should not and will not limit the present invention described in detail in the claims.

[0039] The experimental methods described in the following examples are conventional methods unless otherwise specified; the reagents and materials described are commercially available unless otherwise specified.

[0040] The regional division provided by the present invention helps the park to classify and manage the entire park into sections, and the site classification provided by the present invention provides management ideas for specific enterprises.

[0041] Example 1: A zoning and classification technical method for differentiated management of pollution prevention and control at a chemical park site is implemented using the following steps:

[0042] (1) Construction of basic information database of chemical park sites.

[0043] In order to make the correlation between site pollution characteristics and key environmental factors clearer, relevant information of the survey area was collected, covering the social economy of the entire survey area, the composition of the chemical industry, the geographical location of the park, the site pollution characteristics, meteorological conditions and other information. Using data analysis methods such as Z-score and Min-Max methods, nearly 100,000 core data were screened and condensed. The Kennard-Stone (KS) algorithm was used to divide the chemical park database into a modeling set and a verification set in a ratio of 7:3.

[0044] (2) Fully consider the heterogeneity, complexity and diversity of the pollution environment characteristics of chemical park sites, covering 23 environmental factors in eight categories, including hydrogeology, soil properties, meteorological conditions, industry categories, years of production, socio-economics, regulatory levels, and pollutant properties. Machine learning algorithms such as GBDT, Xgboost, and RF are integrated to train and test models to analyze and identify key elements of chemical parks, and quantitatively evaluate the contribution of various environmental covariates to the depth of pollution impact of chemical park sites.

[0045] In the GBDT model, the number of iterations was set to 100, the learning rate was set to 0.2, the minimum number of samples was set to 5, and other parameters were kept at their default values. In the XGBoost model, the learning_rate was set to 0.01, the gamma was set to 0.1, and other parameters were kept at their default values. In the RF model, the number of trees was set to 1000, the minimum number of samples of the terminal nodes was 5, and other parameters were kept at their default values. During the model training and parameter tuning process, environmental factors with an importance of less than 3% were eliminated, and the model accuracy evaluation results showed that the determination coefficient (R 2 ) gradually tends to 1, and the root mean square error (RMSE) gradually decreases to 0, which verifies the scientificity and effectiveness of environmental factor screening and elimination.

[0046] In addition, during the training process, the environmental covariate feature gain may be negative, which indicates that its contribution to the analysis and identification of chemical park characteristics is small and negligible. Such environmental covariates are directly deleted, and the model is trained again on the screened data set. After three progressive screenings, the environmental factors were reduced from 8 categories to 6 categories, and from 23 items to 14 items. 6 categories and 14 items of site pollution characteristic environmental factors were identified, such as Figure 1 As shown (from top to bottom, there are 23 items in 8 categories, 19 items in 7 categories in the first screening, 16 items in 6 categories in the second screening, and 14 items in 6 categories in the third screening).

[0047] The 6 major environmental factors screened include production years, industry categories, regulatory levels, hydrogeology, pollutant properties, and social economy. The 14 environmental factors include production years, industry categories, surrounding population density, production and operation area, pollutant mobility, pollutant volatility, vadose zone soil medium type, runoff potential, groundwater depth, terrain wetness index, rainfall, soil / groundwater monitoring, air / surface water monitoring, and early warning management. In addition, through a 10-fold cross-validation of the integrated model, the results showed that its prediction accuracy was as high as 97.5%, indicating the superior performance of the integrated model in predicting pollution risks and the key role of the extracted environmental factors in improving prediction accuracy.

[0048] (3) Based on the site pollution environmental factors that have been screened and extracted by the integrated model, the principal component analysis method (PCA) is used to further reduce the dimension and simplify and finally clarify the key factors that characterize the site pollution risk; among them, n_components is set to 12, and other parameters are default values. The analysis results show that the cumulative contribution rate of the five major environmental characteristics of production years, industry category, pollutant properties, hydrogeology and regulatory level is 89% (>85% is a key factor). Figure 2 As shown, years of production, industry category, pollutant mobility, pollutant volatility, soil medium type in the vadose zone, runoff potential, groundwater depth, terrain wetness index, rainfall, soil / groundwater monitoring, air / surface water monitoring, and early warning management (the cumulative contribution rate of these 12 items is 89%, and the key factors that meet the requirement of >85%) are the key factors affecting the pollution of chemical park sites, providing a scientific basis and technical support for analyzing the pollution characteristics of chemical park sites.

[0049] (4) Based on the extracted key elements of environmental pollution in chemical industrial parks, we conducted an analysis based on more than 2,000 actual sampling survey cases collected during the survey, and combined them with the typical characteristic pollution proportions of polluted sites to further quantitatively evaluate their correlation with site pollution characteristics. Figure 3-Figure 5 And as shown in Table 1.

[0050] In terms of production years, we found that there were significant differences in the contamination rates of plots with a production period of ≤3 years, 4-10 years, 11-30 years, and >30 years. Due to long-term production activities and cumulative pollutant emissions, sites with a production period of more than 30 years are more prone to pollution.

[0051] In terms of industry categories, industries that are prone to soil pollution (coking, organic chemical raw material manufacturing, chemical pesticide manufacturing, metal surface treatment and heat treatment processing) have more significant site pollution, while other production and operation industries that may cause soil pollution (textiles, clothing, leather, pigments, coatings, rubber, plastics and their products) have the second highest site pollution rate. The lowest site pollution rate is in industries that do not involve production and operation that may cause soil pollution (other cleaning services, cargo transportation, warehousing, wood processing, etc.).

[0052] In terms of runoff potential, when the runoff potential of the site is greater than 0.4, the surface water is highly mobile, and the probability of pollutants migrating and diffusing with the water flow increases, resulting in an expansion of the pollution range; in particular, when the runoff potential of the site is greater than 0.6, the pollution rate is further increased.

[0053] In terms of terrain wetness index, pollutants enter the groundwater layer through infiltration and runoff. Due to the large catchment area and high-humidity soil environment, the site with a terrain wetness index greater than 7 has a greater amount of pollutant dissolution and diffusion flux, aggravating the degree of pollution.

[0054] In terms of rainfall, there are significant differences in the pollution rates of plots with annual rainfall of <500, 500-900, 901-1200, and >1200 mm. Among them, sites with annual rainfall exceeding 1200 mm are more likely to carry surface pollutants into the ground, increasing the pollution load of soil and groundwater.

[0055] In terms of groundwater depth, there are significant differences in the pollution rates of plots with groundwater GD<3, 3≤GD<10, and GD≥10. The groundwater depth of contaminated sites is mostly concentrated in GD<3m.

[0056] In terms of soil structure of the vadose zone, there are significant differences in the contamination rates among the plots of "fill or thin layer", "gravel, sand and gravel", "silt, soil", and "clay". Among them, the sites with fill or thin layer have loose geological structures, and pollutants are more likely to penetrate and diffuse downward.

[0057] In terms of pollutant mobility and volatility, 95% of land pollution is concentrated in the underground 0-4.5 m. Polycyclic aromatic hydrocarbons and pesticides are mostly concentrated in the soil surface. Organic pollutants such as benzene and chlorinated hydrocarbons are more likely to migrate deep into the soil, increasing the depth of pollution.

[0058] In terms of supervision level, from the perspective of atmospheric surface water monitoring, soil groundwater monitoring, and early warning supervision, the pollution rate is reduced when there is supervision. However, due to the lack of an effective soil and groundwater pollution monitoring and management system in most chemical parks, the pollution problem is more prominent.

[0059] Table 1

[0060]

[0061] (5) After analyzing and quantifying each feature, a scoring system is constructed based on the contribution rate as shown in Table 2.

[0062] Table 2

[0063]

[0064]

[0065] Note: a: When the grade is level Ⅰ, the score is the score of the environmental factor sub-category; except for level Ⅰ, the other grades are obtained by dividing the number of grades equally. For example, there are 4 grades in the production years. First, divide 11 points by 4 to get 2.75. The lowest score is the lowest grade. Level Ⅳ is 2.75×1=2.75 points; Level Ⅲ is 2.75×2=5.5 points; Level Ⅱ is 2.75×3=8.25 points; Level Ⅰ is 2.75×4=11 points; if the pollutant is not volatile, the score of Level IV is 0; if there is monitoring, the score is 0. b: In the pollutant property scoring system, if there are multiple pollutants, the maximum value is used to score the mobility and volatility of the pollutants.

[0066] (6) Based on the big data analysis results of the small- and medium-scale characteristics of the park sites and the needs of different management objects in the park, the method principle of "regional zoning based on pollution status + classification management based on land use scenarios, site classification based on plot risk type + phased policy implementation based on the urgency of governance" was formulated according to different land use scenarios to achieve classified management and phased policy implementation for chemical park areas and sites of different regional scales, industry types, development intensities, management levels, land use scenarios, and pollution levels.

[0067] (7) Regional zoning Based on the level of pollution risk and the interaction between the park and sensitive ecological targets, different areas in the park can be divided into five zones: heavy pollution area, high risk area, medium risk area, low risk area and ecologically sensitive area. Figure 6 As shown. Zoning is a step-by-step judgment method, and the specific process is as follows: ① The first step is the conditional judgment method. If the survey results exceed the control value or the human health risk value, it is a heavily polluted area; ② The second step is the comprehensive scoring method. If the score is >70, it is a high-risk area; ③ The third step is the conditional judgment + comprehensive scoring method. If the survey quality is poor or the site is not surveyed, and the production period exceeds 30 years or an environmental accident has occurred or the pollution risk score is >50, it is a medium-risk area; ④ The fourth step is the conditional judgment method. If there are key river basins or major drinking water sources or ecological protection areas within 1 km of the surrounding area, it is an ecologically sensitive area, otherwise it is a low-risk area.

[0068] After the area is divided, it is classified and managed according to the land use scenarios in each area. For example, if there are areas that cannot be repaired in the heavily polluted area, they are classified as storage control areas, and if they are decommissioned and redeveloped areas, they are classified as restoration and utilization areas, etc. Corresponding management measures are also provided in the technical route.

[0069] (8) Based on the constructed zoning comprehensive scoring system, a database of 1,500 actual surveyed sites was selected to verify the scientific effectiveness of the zoning comprehensive scoring system. The matching degree between the simulation results of the scoring system and the actual investigation and judgment results is affected by differences in environmental characteristics, data sources and model parameter settings. The matching degree of the high-risk area is about 70%. The pollution sources accumulated in long-term production, the multivariate characteristics of high-pollution industries, complex hydrogeological conditions and inadequate supervision lead to insufficient capture of environmental characteristics, which increases the difficulty and uncertainty of the scoring system simulation; the matching degree of the medium-risk area is about 90%. Thanks to the clear production years, industry and pollutant properties, relatively stable hydrogeological conditions and moderate supervision level, it can accurately reflect and simulate the actual situation; the matching degree of the low-risk area is about 80%. There are fewer soil and groundwater pollution problems, low environmental sensitivity and strict supervision measures have reduced the complexity of capturing environmental characteristics, but there may still be environmental blind spots that have not been captured. Overall, the average matching degree between the simulation results of the scoring system and the actual investigation and judgment results is as high as 85%. The constructed regional zoning comprehensive scoring system is reasonable and reliable.

[0070] Example 2: Site classification and phased management method for differentiated management of site pollution prevention and control in chemical park enterprises

[0071] The site classification adopts a step-by-step judgment method, comprehensively considering the site pollution risk, migration risk, potential risk and surrounding risks, and is divided into health prevention and control level, migration prevention and control level, investigation and identification level, protection monitoring level, and low risk level according to the risk information that can be judged on the site. Figure 7 The specific process is as follows: ① If the site pollution exceeds the second-class land use control value or the human health risk value or has a serious odor, it is a health prevention and control level; ② If the pollution exceeds the screening value and the pollution factor mobility M>0.01, or there are coarse sand, pebbles, fine sand, sub-sand, silt, etc. in the natural soil, it is a migration prevention and control level; ③ If the pollution exceeds the screening value or the plot risk concept model is missing or the production period exceeds 30 years or a major environmental pollution accident has occurred within 1 km of the surrounding area, it is an investigation and identification level; ④ If there are key watersheds, ecological protection areas, drinking water sources, etc. within 500 m around the plot, it is a protection monitoring level, otherwise it is a low risk level.

[0072] After the site is graded, the scenarios are classified according to the degree of urgency of governance, and different levels of management and environmental goals in chemical sites are comprehensively considered. Different levels and stages of pollution prevention and control requirements are put forward, including eight categories: initiating governance, initiating control, long-term prevention and control, real-time monitoring, supplementary investigation, initiating detailed adjustment, initiating initial adjustment, and dynamic monitoring. For the health prevention and control level, governance or control is initiated; for the migration prevention and control level, enterprises whose pollution risks have migrated outside the site should be subject to long-term prevention and control, and sites with a migration trend of pollution risks should be subject to real-time monitoring; for the investigation and identification level, sites with a migration trend of pollution risks should be subject to real-time monitoring, sites with unclear migration characteristics of pollution risks should conduct supplementary investigations, and sites with excessive initial adjustments should initiate detailed adjustments; for the protection monitoring level, initial adjustments should be initiated; for the low-risk level, dynamic monitoring is recommended.

[0073] The above-mentioned embodiments only express the preferred implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the attached claims.

Claims

1. A method for differentiated management of pollution prevention and control in chemical industrial parks, characterized in that: include: (1) Combining the influence of eight major environmental factors on site pollution, including hydrogeology, soil properties, meteorological conditions, industry category, years of production, socio-economics, regulatory level, and pollutant properties, the influence factors of differentiated management of pollution prevention and control in chemical parks were determined by integrating GradientBoosting Decision Tree, Extreme Gradient Boosting, and Random Forest for model training and testing analysis, and the contribution of each influencing factor to site pollution in chemical parks was quantitatively evaluated. (2) From the identified influencing factors of differentiated management of pollution prevention and control at chemical industrial parks, reduce the dimension and identify the environmental characteristic indicator factors of differentiated management of pollution prevention and control at chemical industrial parks; (3) Based on the quantitative results of environmental characteristic index factors at the chemical park site and combined with the contribution rate, a scoring system is constructed to divide the level of each environmental characteristic index factor and assign points to each level; (4) Based on the graded results of environmental characteristic index factors, combined with any one or more of the survey risk, regional risk, historical risk and surrounding risk of the chemical park site, a technical route for regional zoning of pollution prevention and control of chemical park sites is established to carry out differentiated management; In step (1), the eight categories of environmental factors include 23 environmental factors, specifically: hydrogeology includes the type of soil medium in the vadose zone, runoff potential, groundwater depth, terrain moisture index, rainfall, elevation, and terrain slope; soil properties include soil bulk density, soil temperature, and soil thickness; meteorological conditions include PM2.5 concentration, PM10 concentration, and average relative humidity of the air; industry category; production years; socio-economics includes surrounding population density and production and operation area; regulatory level includes soil / groundwater monitoring, air / surface water monitoring, early warning management, and emergency management; pollutant properties include pollutant mobility and pollutant volatility; In step (2), the dimensionality reduction identification method is the principal component analysis method; the environmental characteristic index factor is obtained by screening the influencing factors with a cumulative contribution rate greater than 85%; In step (3), the level of each environmental characteristic index factor is divided according to the pollution rate of the environmental characteristic index factor in the chemical park site; and the contribution rate of the pollution of the chemical park site is assigned points according to the environmental characteristic index factor; In step (3), the scoring system includes three levels: major environmental factor categories, minor environmental factor categories and environmental factor indicator layers, wherein the minor environmental factor categories are the environmental characteristic indicator factors; the minor environmental factor categories are sub-levels of the major environmental factor categories, and the environmental factor indicator layer is a sub-level of the minor environmental factor categories; the major environmental factor categories include production years, industry categories, hydrogeology, pollutant properties, and regulatory levels; the minor environmental factor categories include vadose zone soil medium types, runoff potential, terrain moisture index, groundwater depth, rainfall, pollutant mobility, pollutant volatility, soil / groundwater monitoring, air / surface water monitoring, and early warning management; after the system is established, it is quantified, firstly, the characteristic importance of the minor environmental factor categories is measured and scored, and the sum is obtained to obtain the major environmental factor category scores; then, combined with the scores specified for the minor environmental factor categories, the scores of the environmental factor indicator layer are divided equally by the number of levels, and the scores of the environmental factor indicator layer are calculated.

2. The method according to claim 1, characterized in that In step (4), the regional zoning adopts a step-by-step judgment method, including two forms: conditional judgment and comprehensive scoring; the conditional judgment is to judge any one or more of the investigation risk, regional risk, historical risk and surrounding risk of the chemical park site; the comprehensive scoring is the scoring system described in step (3).

3. The method according to claim 2, characterized in that When the survey results exceed the control value or exceed the human health risk value, the area of ​​the chemical park is a heavily polluted area; when the total score calculated according to the scoring system is greater than 70 points, the area of ​​the chemical park is a high-risk area; when the survey quality of the area is poor or has not been surveyed, and the total score calculated according to the scoring system is 50<≤70, or the production period of the chemical park site is greater than 30 years, or an environmental accident has occurred, the area of ​​the chemical park is a medium-risk area; when the total score calculated according to the scoring system is ≤50, and there are no key river basins, major drinking water sources, or ecological protection areas within 1 km of the surrounding area, the area of ​​the chemical park is a low-risk area.

4. The method according to claim 3, characterized in that When the total score calculated according to the scoring system is ≤50, and there are key river basins, major drinking water sources, or ecological protection areas within 1 km of the surrounding area, the chemical park area is an ecologically sensitive area.

5. The method according to claim 4, characterized in that Based on the results of regional zoning and combined with the utilization scenarios of different areas in the chemical park, classified management is carried out.

6. The method according to claim 5, characterized in that Heavily polluted areas that cannot be repaired are classified under the storage control category; heavily polluted areas that are retired and redeveloped are classified under the restoration and utilization category; heavily polluted areas that are retired and idle, under production control, and high-risk areas with pollution are classified under the pollution control category; high-risk areas without pollution, low-risk areas that have been developed and utilized, and ecologically sensitive areas that have been developed and utilized are classified under the pollution prevention category; the medium-risk areas, low-risk areas that have not been developed and utilized, and ecologically sensitive areas that have been developed and utilized are classified under the pollution identification category.

Images