A Classification System for the Sustainable Management of Contaminated Sites Coupled with Risk Identification and Value Accounting

Currently, site contamination is considered to be a sustained, international environmental challenge, and there is an urgent practical need to build a core theoretical system and technical methodology for the sustainable risk management of soil contamination, together with its prevention and control. We aim to improve the risk management of contaminated sites in the post-remediation era, in line with the current trend of sustainable development. The work is based on the theory of sustainability science and the eco-environmental zoning system., In this study, we build a conceptual model that can be used to classify the sustainable performance of contaminated sites in terms of risk management in line with the existing environmental management system for contaminated sites in China. To provide a scientific decision-making basis and technical support for the refined classification management of soil environments in China during the 14th Five-Year Plan period, five typical contaminated sites were selected for a quantitative evaluation by applying multi-technical approaches, including sociological, economic and statistical methods. The results showed that the sustainable performance of contaminated sites with regard to management was affected not only by pollution risk factors but also by potential utility benefits. Specified management strategies should be developed according to different levels of sustainability so as to achieve the goals of improving land use efficiency and enhancing urban functions.


Introduction
With the adjustment and transformation of industrial structures and the development of urbanization, many contaminated sites located on industrial land have caught stakeholders' attention concerning the issues regarding their cleanup and redevelopment. Contaminated sites (also referred to as brownfield sites) are sites with pollution hazards beyond the acceptable risk levels for human or ecosystem health [1][2][3]. The implementation of contaminated site remediation has been practiced in developed countries for nearly half a century and evolved through a gradual change in management mindsets from the complete removal of contaminants to risk-based management and, finally, to green and sustainable risk management, supporting the wide-ranging consideration of risk management technology in the social, environmental and economic dimensions. As a result, in developed countries, organizations including ASTM, USEPA, ISO, CLARINET and SURFs have generated active responses by issuing a series of practice frameworks, standard guidelines and technical assessment guidelines to develop a systematic and comprehensive sustainable risk management system for contaminated sites [4][5][6][7][8]. The challenge is that the contradiction between the large number of sites and the shortage of resources, as well as the high cost of remediation and the urgent need for the elimination Republic into five redevelopment priority levels, namely lower, low, medium, high and higher, in terms of the three dimensions of regional redevelopment potential, i.e., site attractiveness, market competitiveness and environmental risk elimination. Moreover, the spatial layout of infrastructures constraining land reuse planning, such as transportation, medical care and education, were taken into account to build a multi-objective decision-making model for optimal land planning [31][32][33][34][35][36][37].
Generally speaking, the sustainable management of contaminated sites is a complicated decision-making process. From a whole-life-cycle perspective, the early risk management strategies mainly focus on cost reduction and the minimization of adverse environmental impacts, while the post-remediation management strategies place greater emphasis on land benefits and social sustainability [38]. However, currently, studies on the risk classification of contaminated sites do not consider the interaction between potential risks and probable land values, which may lead to a biased decision-making process for site risk management, without considering the supporting roles of other decision dimensions (environment, society and economy). Therefore, to reveal the mechanism of interaction between risk management and regional sustainable development of contaminated sites, there is an urgent need to establish a decision-making support system for contaminated site management classification. With the aim of filling certain gaps in previous research and guide site sustainability management, this work is structured as follows: (1) The methodology used to develop an indicator system, process data, and quantify the evaluation indicators is presented; (2) Using the results, first, we conduct a detailed analysis of the risk characterization and land economic values of the selected contaminated sites, and then the category framework for the contaminated sites is mapped to explain the important factors that can significantly influence the sustainability levels and support of future site management decisions; (3) In the last section, the main conclusions are highlighted, and challenges to be addressed in future work are proposed.

Subsection
This study was conducted in five typical sites in China, i.e., the Xinguang Stainless Steel Factory (XSSF), the former Dongfang Chemical Plant (DCP), the Nanning Chemical Plant (NCP), Shenyang Chemical Plant (SCP) and the former Great Wall Chemical Plant (GWCP). The information collected on the sites included investigation and remediation reports, official government documents, yearbooks, questionnaire surveys and related websites. The location, basic information and risk management factors of the sites are shown in Figure 1 and Table 1.

Assessment Indicator System
The classification indicator system for contaminated site risk management was designed by combining sustainable development theory with a risk factor and cost-benefit analysis, including two groups: the risk index and value index. The risk index was used to evaluate the existing contamination levels, the vulnerability of the risk receptors and potential risks caused by operational enterprises., while the value index was applied to analyze the net benefit of site management in terms of the site remediation cost and the economic value of the land's reuse. The definitions of the assessment indices are shown in Table 2.

Assessment Indicator System
The classification indicator system for contaminated site risk management was designed by combining sustainable development theory with a risk factor and cost-benefit analysis, including two groups: the risk index and value index. The risk index was used to evaluate the existing contamination levels, the vulnerability of the risk receptors and potential risks caused by operational enterprises., while the value index was applied to analyze the net benefit of site management in terms of the site remediation cost and the economic value of the land's reuse. The definitions of the assessment indices are shown in Table 2.

The 1st-Level Index
The 2nd-Level Index The 3rd-Level Index Definition

Risk index
Vulnerability of risk receptors Population density Population located within 1 km of the site.

Distribution of sensitive objectives
The number of land use scenarios within 1 km of the site: primary and secondary schools, medical and health care and social welfare facilities.

Ecological function areas
The number of ecological function areas (listed in spatial planning reports) within 1 km of the site: water conservation, biodiversity protection, flood regulation and the supply of agricultural products.

Potential contamination risk
Number of operational enterprises The number of operational enterprises within 1 km of the site: non-ferrous metal mining and processing, petroleum processing, chemical, coking, electroplating, etc., or those that have caused environmental pollution in recent years.

Soil contamination
Number of contaminants in soil exceeding the relative standards, as recorded in remediation reports.

Groundwater contamination
Number of contaminants in groundwater exceeding the relative standards, as recorded in remediation reports.

Benefit index
Economic loss Remediation cost The cost required to remove existing contaminants. Economic value

Regional economic development
The edge effects that could be produced through surrounding environmental improvements, i.e., house appreciation.

Land price
The transfer price of the land after remediation.

Employment opportunities
The value of the job opportunities that could be provided by redeveloping a remediated site for a commercial purpose or into a park.

Recreation and leisure
The value of the recreation and leisure services that could be provided by redeveloping a remediated site into a park.

Ecological value
Regulating services Mainly includes gas regulation, climate regulation, environmental purification and hydrological regulation.
Supporting services Mainly includes soil conservation, the maintenance of nutrient cycling and biodiversity.

Data Collection and Accounting Methodology
Regarding the risk index, the calculation of the population density is based on the Resource and Environmental Science and Data Center (https://www.resdc.cn/, accessed on 10 August 2022). Using the Zonal Statistics as Table tool in ArcGIS, the raw data of the contamination level indicator are obtained directly from the site risk assessment report, and other risk indices are measured by collecting POI data from the Gaode open platform.
For the benefit index, the remediation cost can be extracted from the bidding documents of remediation projects, the land transfer price can be obtained from the Chinese land marketing website (https://www.landchina.com, accessed on 10 August 2022), and other benefit indices can be calculated with equations as follows below.
Regional economic development: A remediated site can not only eliminate health and ecological risks but also deliver social and economic benefits, such as the promotion of social stability and regional economic growth. In this study, the stigma effect of the contaminated site on the adjacent properties is predicted in order to evaluate the approximate economic development potentials that could be obtained through site management. The calculation of the stigma or rebound effect is as shown in the following equation: where Den is the population density within 1 km of the studied site (person/km 2 ), A 1km is the land area within 1 km of the studied site (km 2 ), FS is the floor space per capita (m 2 /person), P is the average house price within 1 km of the studied site (CNY/m 2 ), and p is the house appreciation ratio (%). Job opportunity: The dominant functions of the redevelopment of contaminated sites into commercial lands are the employment opportunities and the resulting increase in personal incomes. In this method, the market value approach is applied to evaluate the income increase by determining the average salaries of employees in various industries. Since this study focuses on urban commercial land, industries such as agriculture, forestry, animal husbandry and fisheries, mining, manufacturing, construction, transportation, storage and postal services are not included in the study boundary. The calculation of job opportunities is as shown in the following equation: where PL i is the average salary of industry i (10,000 CNY/person), and SAL i is the number of employees in the industry i.
Recreation and leisure: The recreation and leisure value represents the revenue generated by redeveloping contaminated sites into open green space, i.e., a park. Generally, such a value includes travel fees and consumer surplus, calculated indirectly by questionnaire surveys among park visitors. The calculation of the recreation and leisure value is as shown in the following equation: where TC is the travel cost (CNY), CS is the consumer surplus (CNY), n is the total number of questionnaire respondents, and N is the total number of park visitors in one year. The ecological value (including regulating services and supporting services): The change in land use can cause significant interactions between ecosystem services (ES) [39]. The major ES provided by contaminated site greening can be defined as regulating services (including gas regulation, climate regulation, environmental purification and hydrological regulation) and supporting services (including soil conservation, the maintenance of nutrient cycling and biodiversity) [40]. Based on land use/land cover (LULC), in this study, we estimated the ecosystem service values (ESV) based on the net primary productivity of the vegetation in the regenerated sites, using the method developed by Xie et al. [41]. The main LULC types involved in the ESV evaluation include mixed broadleaf-conifer forests, broad-leaved forests, shrubs, grasslands, scrubs, artificial lakes and wetlands. The calculation of the ESV is as shown in the following equation: where A is the area of the evaluated ecosystem (m 2 ), F ni is the coefficient of the area values per unit of ES n in area i, and D is a standard equivalent factor of the ESV (CNY/m 2 ). Considering the fact that the soil conservation service is affected by multi-complex factors and that soil conservation simulation data are almost entirely unavailable, the soil conservation service is not corrected in this study, while the area value per unit equivalent factors of other ES are adjusted by the NDVI (normalized difference vegetation index), or precipitation factor, in order to conform to the environmental situation of the studied site. The calculation of F ni is as shown in the following equation: where B i is the average NDVI of the evaluated ecosystem in area i in the last ten years, B is the average NDVI of the evaluated ecosystem in China in the last ten years, W i is the average precipitation in area i in the last ten years (mm·yr −1 ), W is the average precipitation in China in the last ten years (mm·yr −1 ), F in1 represents ES including gas regulation, climate regulation, environmental purification, the maintenance of nutrient cycling and biodiversity, F in2 represents hydrological regulation services, and F in3 represents soil conservation services. The net profit of crop production per unit area in the farmland ecosystem is used to measure a standard equivalent factor of the ESV [42]. The calculation of D is shown in the following equation: where m i is the area of crop i (hm 2 ), p i is the average price of crop i (CNY/ton), q i is the yield per unit area of crop i (ton/hm 2 ), and M is the total area of all the crops (hm 2 ). Normalization of the index: Because the unit and numerical levels of an index are different, the raw data should be standardized to enable a comparative evaluation of different sites. The normalization of data is as the following equation: where x is the initial value of the evaluation indicator, max represents the maximum value, and min represents the minimum value.

Classification Method
Based on Equations (1)- (7), the standardized values of the risk and benefit indices could be weighted to obtain an overall risk value and benefit value ranging from 0 to 1. Then, the risk value was equally divided into three levels, i.e., low-risk (0-0.3), medium-risk (0.3-0.6) and high-risk (0.6-1). Similarly, the benefit value was classified into three levels, i.e., low-benefit (0-0.3), medium-benefit (0.3-0.6) and high-benefit (0.6-1) as shown in Figure 2. As a result, the sustainability performance of contaminated site risk management was categorized into three levels depending on both the risk value and benefit value, i.e., high-sustainability, medium-sustainability and low-sustainability, which are expected to provide support for scientific decision-making regarding future site managemen.

Risk Assessment of the Contaminated Sites
In terms of the vulnerability of risk receptors, potential contamination risk and current contamination level, the risk identification results of the five contaminated sites are shown in Table 3. The DCP, NCP and SCP sites are adjacent to the city center, meaning a greater population density (>1000), more sensitive objectives (>10) and more severe contamination (both soil and groundwater contamination) than those located in the suburbs, i.e., XSSF and GWCP, which demonstrates the higher risk posed by site contamination. Except for DCP, no contaminants are likely to be an input for the other four sites. Table 3. The results of the risk identification analysis of the contaminated sites.

Risk Assessment of the Contaminated Sites
In terms of the vulnerability of risk receptors, potential contamination risk and current contamination level, the risk identification results of the five contaminated sites are shown in Table 3. The DCP, NCP and SCP sites are adjacent to the city center, meaning a greater population density (>1000), more sensitive objectives (>10) and more severe contamination (both soil and groundwater contamination) than those located in the suburbs, i.e., XSSF and GWCP, which demonstrates the higher risk posed by site contamination. Except for DCP, no contaminants are likely to be an input for the other four sites.

Benefit Assessment of the Contaminated Sites
The sustainability assessment of the economy is mainly based on the cost input and benefit output during the management of sites throughout their life cycles. In terms of economic loss, economic value, social value and ecological value, the value accounting results of the five contaminated sites are shown in Figure 3.

5.4/m 2 .
Based on face-to-face field interviews, the recreational benefits of the DCP were evaluated using Equation (3). The questionnaire survey was conducted on 2 April 2022 at the DCP site, and 140 questionnaires (out of a total of 150) were collected from the park visitors. No individual information was publicized in our study, and all the interviewees signed an "Informed Consent Form" before our investigation. Due to the limitation of the length of the paper, detailed information, including the questionnaire design, collected data and data analysis, are not provided here but can be accessed upon request to the authors.

Benefit Assessment of the Contaminated Sites
Using Equation (7), a numerical value for a particular indicator was normalized against the maximum or minimum value to transform it into a score ranging from 0 to 1. As shown in Table 4, an indicator weight was assigned based on the hypothesis that each indicator belonging to the same criteria contributed equally to the site's sustainability performance. For example, here, six risk indices contribute 1/6 (15~17%) to the total risk score. The remediation actions and types of land reuse have important impacts on land values. As illustrated in Figure 3, although the investment for the remediation of the SCP site is significantly higher than that of the other sites, its redevelopment produces the highest direct land value, amounting to CNY 9552.91/m 2 . The assessment results show that the XSSF site has the highest social value, with CNY 8486.84/m 2 , due to its tourism and sightseeing functions, while the ecological value contributes least to the land value compared with the economic value and social value; NCP has the highest ecological value, with CNY 5.4/m 2 .
Based on face-to-face field interviews, the recreational benefits of the DCP were evaluated using Equation (3). The questionnaire survey was conducted on 2 April 2022 at the DCP site, and 140 questionnaires (out of a total of 150) were collected from the park visitors. No individual information was publicized in our study, and all the interviewees signed an "Informed Consent Form" before our investigation. Due to the limitation of the length of the paper, detailed information, including the questionnaire design, collected data and data analysis, are not provided here but can be accessed upon request to the authors.

Benefit Assessment of the Contaminated Sites
Using Equation (7), a numerical value for a particular indicator was normalized against the maximum or minimum value to transform it into a score ranging from 0 to 1. As shown in Table 4, an indicator weight was assigned based on the hypothesis that each indicator belonging to the same criteria contributed equally to the site's sustainability performance. For example, here, six risk indices contribute 1/6 (15~17%) to the total risk score.
The overall risk value, benefit value, rankings and classification of the individual sites are displayed in Figure 4. We can see that the classification can be equal between sites ranking differently in terms of their sustainability performance. As demonstrated below, the risk management actions of the XSSF, NCP and SCP sites are classified as high sustainability, while the sustainability performance of the DCP and GWCP is relatively low. It is clear that the classification results depend on both the risk characteristics and the benefit feedback. The overall risk value, benefit value, rankings and classification of the individual sites are displayed in Figure 4. We can see that the classification can be equal between sites ranking differently in terms of their sustainability performance. As demonstrated below, the risk management actions of the XSSF, NCP and SCP sites are classified as high sustainability, while the sustainability performance of the DCP and GWCP is relatively low. It is clear that the classification results depend on both the risk characteristics and the benefit feedback. The risk value assessment results in Figure 4 show that the contamination risks of the NCP and SCP sites are higher than those of the other three sites due to the fact that sensitive objectives are crowded around the two chemical plants, especially primary and secondary schools, which number as high as 55 and 43, respectively. Therefore, the risk receptors around the NCP site and SCP site show high vulnerability, and the areas impacted by contamination risks are enlarged due to local overpopulation. Comparatively, the small population and number of sensitive objectives around XSSF mean that it has the lowest risk value among all the sites. In addition, the unacceptable human health risk posed by the groundwater in the SCP, NCP and DCP sites results in significant differences The risk value assessment results in Figure 4 show that the contamination risks of the NCP and SCP sites are higher than those of the other three sites due to the fact that sensitive objectives are crowded around the two chemical plants, especially primary and secondary schools, which number as high as 55 and 43, respectively. Therefore, the risk receptors around the NCP site and SCP site show high vulnerability, and the areas impacted by contamination risks are enlarged due to local overpopulation. Comparatively, the small population and number of sensitive objectives around XSSF mean that it has the lowest risk value among all the sites. In addition, the unacceptable human health risk posed by the groundwater in the SCP, NCP and DCP sites results in significant differences in the risk value, indicating that contaminants in the groundwater tend to contribute more than those in the soil to risk formation and transmission.
In terms of the benefit value assessment results in Figure 4, the net benefits of the remediation actions of the SCP, XSSF and NCP sites are significantly higher than those of DCP and GWCP. Specifically, the redevelopment of the SCP site into residential land generates immense appreciation effects on the remediated site and surrounding properties (CNY 9073.29/m 2 ), though this also results in the highest remediation cost per unit area (CNY 480.33/m 2 ). In the case of XSSF, the job opportunities and tourism revenue form a greater proportion of the total value due to the site's particular location in a tourism city (Huangshan City). Meanwhile, for the NCP site, located in the city center, the high land transfer price and low remediation cost (CNY 35.13/m 2 ) enable a relatively higher net benefit (CNY 7094.55/m 2 ). Currently, there is no reuse plan for the GWCP site; however, considering its geographical location and risk management countermeasures, an industrial purpose may be its ideal reuse scenario and could produce social benefits by providing job opportunities. Although the DCP site has great property appreciation (CNY 5.39 billion), the park reuse scenario limits its market value and economic output, thus resulting in the lowest net benefit (CNY 768.38/m 2 ).

Conclusions
This study established a value-based approach to improve the existing risk classification system for sustainable site regeneration, through which contaminated sites with high pollution risks and high reuse values can be identified in order to implement risk control activities and subsequent development planning. In particular, for regions with numerous contaminated sites subjected to financial challenges and land shortages, the novel classification system has a wider range of applications; not only could it eliminate the pollution risks, but it also could maximize the land utilization profits. The main conclusions regarding the case study of the five contaminated sites are as follows: (1) The size of the population and the number of sensitive objectives have a major impact on the contamination risk levels. At present, the human health risk is the primary concern of the risk assessment of contaminated sites. Contaminated media can lead to significant risk differences; groundwater pollution has greater effects in changing the risk values compared with soil pollution. (2) The overall value of a site depends on its economic value, social value and ecological value. Among these factors, the appreciation potential of the site and surrounding properties as a result of site remediation actions plays a vital role in increasing the overall value of the site. The value produced by providing job opportunities is an important aspect of the social value, and the ecosystem service value accounts for a relatively small proportion of the total value. (3) The sites should be treated differently according to the corresponding risk and benefit evaluation results. For sites with high sustainability, decisions regarding their prior risk management and reuse should be made as soon as possible to eliminate health risks and generate substantial profitability through land reuse. For sites with a medium level of sustainability, a restricted development strategy is recommended, and management patterns with higher returns should be actively explored for the purpose of eliminating existing risks. For sites with low sustainability, management decisions should focus on preventing contamination diffusion rather than rapid redevelopment while simultaneously conducting a scenario analysis to maximize the land values.
Compared with previous studies, our research differs in two aspects: Firstly, the risk evaluation depended on indicators measuring the existing risks and the potential risks posed by local pollution sources, which enabled a dynamic simulation according to the spatial emission characteristics of the contaminants. Secondly, by focusing on the current challenges involved in sustainability assessments for brownfield regeneration, this paper proposed a quantitative method that can be used to monetize the social, economic and environmental benefits of site management, which performs better than subjective approaches based on expert scoring or descriptive arguments. However, several challenges remain to be addressed in further research. For example, the evaluation pattern is strongly dependent on the indicators used, but here, further analysis of the possibly more influential indicators was not discussed in depth due to the scarcity of data. Additionally, the weights of both the overall index and sub-indices are hypothesized to be equal according to the calculation process, although they actually have different contributions to sustainability performance. Moreover, the indicator of the potential risks is measured by the input flux of contaminants rather than the number of surrounding enterprises.
Author Contributions: Conceptualization, X.L. and W.C.; methodology, X.L. and W.C.; validation, X.L.; formal analysis, S.Y.; investigation, X.L. and S.Y.; resources, X.L. and W.C.; data curation, X.L. and S.Y.; writing-original draft preparation, S.Y.; writing-review and editing, X.L.; visualization, S.Y.; supervision, X.L. and W.C.; project administration, X.L. and W.C.; funding acquisition, X.L. All authors have read and agreed to the published version of the manuscript. Data Availability Statement: Even though the data is not publicly available it can be requested to the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.