Environmental and socio-economic sustainability assessment of remediation alternatives for a contaminated oil refinery site in Southern China

Environmental and socio-economic sustainability assessment of remediation alternatives for a contaminated oil refinery site in Southern China | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Environmental and socio-economic sustainability assessment of remediation alternatives for a contaminated oil refinery site in Southern China Yingluo Jia, Xianglan Li, Hongzhen Zhang, Chunlong Zhang, Yafei Wang, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6889334/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 Nov, 2025 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Refineries are an important industrial legacy, emitting large amounts of petroleum hydrocarbons and highlighting the need for the green and sustainable remediation of contaminated oil refinery sites globally. This study aimed to evaluate the environmental and socio-economic sustainability of thermal desorption (TD) and stabilization/solidification (S/S) strategies for remediating a contaminated oil refinery site in Sichuan Province, China. We calculated the energy consumption, greenhouse gas emissions, and air pollutants across different remediation phases, and quantified the sustainability of the remediation using multi criteria analysis (MCA) and Bayesian networks (BNs). TD consumed 518,088,020.45 million British thermal units (MMBTU) of energy and emitted 8,915.55 kgCO 2 -eq carbon, compared with 555,706.31 MMBTU and 5,102.95 kgCO 2 -eq, respectively, for S/S. Socio-economic appraisal showed that S/S was associated with lower economic costs, better worker safety, and greater sustainability, while TD provided more employment opportunities. BN analysis further predicted a 71% probability of high social benefits and a 56% probability of economic benefits from converting the site into a public park. These results highlight the need for integrated strategies balancing environmental remediation, economic viability, and community engagement, to provide a framework for the sustainable urban regeneration of industrial legacies. Earth and environmental sciences/Environmental sciences Earth and environmental sciences/Environmental social sciences Contaminated site remediation Multi-criteria assessment Decision-making Oil refinery site Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1 Introduction The contamination of soils by oil and petroleum-based hydrocarbons as a result of accidental spills is causing increased environmental concern 1 . In China, 77% of large-scale contaminated sites are polluted by organic contaminants including petroleum hydrocarbons 2 . These pollutants have adverse impacts on human health and ecosystems, and the remediation of former refinery sites has attracted widespread attention 3 . Among the diverse remediation technologies, thermal desorption (TD) is noted for its high efficiency and broad applicability 4 , while stabilization/solidification (S/S) is a cost-effective and robust method for immobilizing contaminants 5 . Remediation professionals face many challenges in terms of incorporating sustainable remediation principles into real-world practice, especially in countries like China where the remediation industry is still in its infancy 6 . Selecting the optimal strategy requires a sustainable remediation framework that balances environmental, economic, and social aspects 7 . While assessment frameworks exist, economic and social impacts are often underrepresented compared to environmental metrics 8 . Specialized tools like SEFA have advanced the quantification of environmental footprints, such as greenhouse gas emissions 9 , and have been applied to petroleum-contaminated sites 10 . However, there remains a lack of comprehensive, multi-criteria sustainability assessments for contaminated oil refinery sites in China, and existing appraisals can be subjective 10 . The present study selected a former refinery in Sichuan Province covering 52.32 hm 2 , associated with severe contamination of the soil and groundwater by petroleum hydrocarbons and heavy metals, necessitating efficient remediation to mitigate environmental and health risks. This study implemented TD and S/S as remediation strategies, and selected soil-remediation data samples for analysis. The objectives of this study were to: (1) conduct a specific and detailed quantitative environmental impact assessment of the remediation alternatives for a contaminated refinery site; (2) develop social and economic indicators and conduct a comparative socio-economic sustainability assessment of TD and S/S using MCA; and (3) conduct a probabilistic analysis of the causal relationship between public acceptance and reconstruction benefits by Bayesian networks (BNs). 2 Materials and methods 2.1 Case study description The refinery (total area approximately 52.32 hm 2 ) was located in Sichuan Province. It had been a significant environmental pollution concern to the locals and a demolition decision was made in 2013 to allow land redevelopment; a new park will be established on the site after appropriate remediation. The remediation scope is shown in Fig. 1 . The whole project needs to remediate 51,153.7 m 3 of soil. The main soils in the restoration area were miscellaneous fill soils, pebbles, and silt clay, and the permeability was poor. The thickness of the silt clay layer was 5.2–13.3 m, with an average thickness of 8.4 m, and the maximum exposure depth was 9.1–16.1 m, with an average depth of 13.42 m. According to the investigation results, the plot was polluted with aliphatic hydrocarbons (C 10 –C 40 ), benzo[a]pyrene, benzo[a]anthracene, dibenz[a,h]anthracene, and the heavy metals cobalt, lead, and arsenic. Based on the formation conditions, the characteristics of the pollutants and the spatial distribution of the remediation area, TD and S/S restoration models were adopted. The environmental footprint-accounting boundary included the entire process of TD and S/S from the beginning of remediation to the end, including construction-site layout, soil remediation, groundwater remediation, and wastewater disposal (Fig. 2 ). The environmental footprint was generated by the use and transportation of fuel oil, electrical machinery, and construction materials and chemicals. SEFA 3.0 (EPA, USA) is a GSR framework decision-support tool that quantifies energy use, GHG emissions, and air pollutants across remediation phases to identify high-impact activities. It also provides qualitative descriptions of the impacts of remediation activities on the land and ecosystems 11 . Additional activities, such as field experiments and MCA, can be conducted to supplement the assessment results. SEFA integrates energy, air emissions, water, and materials using lifecycle inventory data to model site-specific scenarios. In this study, TD and S/S strategies were evaluated during the remediation process using SEFA 3.0, including analyzing inputs such as electricity, fuel, and materials to generate outputs for energy consumption, GHG emissions, and air pollutants, including nitrogen oxide (NO X ), sulfur oxide (SO X ), particulate matter with a diameter ≤ 10 µm (PM10), and hazardous air pollutants (HAPs). 2.2 Remediation strategies Two major site-remediation technologies, TD and S/S, were used. TD is a process that heats the contaminated soil to volatilize pollutants and then collects them for treatment, ensuring effective remediation 1 . The equipment produces high-temperature flue gas through gas combustion on the outer wall of the TD chamber, transfers the heat to the material in the chamber, and achieves soil heating 12 up to 800℃, at which almost all the soil can be vaporized and separated. The high-temperature exhaust gas from TD is treated successively and then introduced into the combustion area of the burner for disposal. S/S removes the contaminated soil and fixes the contaminant by adding a stabilizer to reduce the risk of migration 13 . TD has no need to excavate and is thus a flexible process with good mobility and can be reused. S/S employs agents such as silicate cement to encapsulate heavy metals, and once the curing process is complete and the concentrations of heavy metals in both acid and water leachates are shown to be below the stipulated standards, the treated material can then be disposed of through in situ or barrier containment and landfilling. The environmental footprints of the two remediation technologies were calculated separately and the sources of the environmental footprints were divided into three scopes: scope 1 involves the direct emission of on-site restoration activities; scope 2 involves the electricity generation; and scope 3 involves other off-site emissions generated from the production, disposal, and transportation of upstream and downstream products and wastes. 2.3 Social and economic sustainability assessment To assess the sustainability of the two major site-remediation technologies, we selected a set of indicators to assess the economic and social sustainabilities. The indicator categories included economic costs and benefits, worker health and safety, communities and community involvement, and public acceptance. Economic indicators were represented by ECON 1–ECON 5 and Social indicators were represented by SOC 1–SOC 5 (Table S1 ). All the indicators were derived from existing guidelines and realistic concerns about the context of the remediation industry in China. The different options were assigned a ranked score of 1–5 framed against an idealized scenario, with a lower score indicating a more-sustainable option and a higher score indicating a less-sustainable option. Multi-criteria models can help remediation practitioners to evaluate multiple conflicting criteria in order to make sustainable remediation decisions about contaminated sites 14 . MCA is a method that is increasingly used for remediation sustainability assessments, to support decision-making. In this study, we used a similar approach to that described by Postle et al. 15 . Both quantitative and qualitative indicators were used to inform the MCA (Table S1 ), and qualitative indicators were assigned a score of 1–5, with lower scores indicating superior outcomes: where is MCA score, is the input value for each indicator, and is the priority weighting for each indicator. The indicators were normalized by weighing the input values against the maximum values across all treatment options and multiplied to provide a score of 0–100. The priority weighting system was based on the concerns of stakeholders. Social impacts were given emphasis in the sustainability assessment. Worker safety was identified as a fundamental social indicator, in light of its regulatory enforcement in many countries, where occupational health standards are increasingly stringent 7 . Public acceptance, measured through community involvement and satisfaction, reflects stakeholder engagement; however, community participation in China is often limited, unless required by environmental impact assessments 6 . Equality was an optional indicator to address concerns of environmental injustice, which are particularly relevant in China where disparities in pollution exposure and remediation benefits have been documented 16 . Local community impacts, such as dust, noise, and traffic congestion, were also considered, with environmental supervision ensuring compliance with national standards, as well as increased traffic volume associated with off-site impacts, including harmful air emissions, noise pollution, and road degradation 17 . Economic impacts were evaluated by balancing costs and benefits. Direct benefits included increased land value, while indirect benefits encompassed employment and local business opportunities 18 . The economic benefits included not only financial expenditures, but also temporal and technological risks. Shorter remediation timelines are often prioritized to accelerate redevelopment and maximize economic returns, reflecting the rapid pace of urban regeneration 19 . To enhance its applicability, the indicator set was structured into core and optional elements, allowing flexibility to adapt to diverse project contexts and stakeholder priorities 20 . 2.4 Overall sustainability appraisal A BN is a graphical probabilistic model used to represent and manipulate uncertain knowledge, and is employed to examine interactions and probability dependencies between variables 21 . It consists of a set of random variables and their conditional dependencies, which are usually represented by a directed acyclic graph. BNs are often constructed to reflect causality and follow a directed path from cause to effect 22 . The goal of BN structure learning is to construct a network using expert knowledge or observational datasets, so that the learned network can maximize the expression of the complex associations between random variables 23 . BNs include parent and child nodes, with each node representing a random variable: parent nodes represent input parameters and child nodes receive inputs from one or more parent nodes, while the directed edges indicate conditional dependencies between the variables. The graphical engine and concept mapping style creates a transparent causal model that can be evaluated with data and expert knowledge at all phases of the model-building process 22 . BNs incorporate the deterministic and stochastic aspects of complex systems, explicitly consider uncertainty in the model inputs, and provide probabilistic predictions with measures of the importance of the input variables (sensitivity analysis). Input parameters are represented as probability distributions, which are derived directly from monitoring data. Parent, child, and endpoint nodes are discretized into ranked states, which allows the evaluation of the combined effects of multiple stressors, including categorical factors or factors with varying units of measurement 24 . In this study, BNs provided a framework to assess citizens’ expectations and opinions about the remediation and transformation of the oil refinery. A structured questionnaire was developed to investigate public expectations and perceptions regarding the remediation and redevelopment of the contaminated oil refinery site. A total of 40 participants were recruited, targeting residents of diverse backgrounds, including local residents, employees of nearby industries, environmental professionals, and students, to capture societal perspectives. The questionnaires were distributed via community events, email, and social media platforms such as WeChat, to ensure broad demographic coverage and minimize selection bias. The study protocol was approved by the Faculty of Geographical Science, Beijing Normal University. In accordance with the regulations of Beijing Normal University, this study is classified as a routine assessment project and, therefore, does not require approval from an Ethics Committee or Institutional Review Board. The study does not involve animal or human clinical trials and is not unethical. The research was conducted in line with the ethical principles outlined in the Declaration of Helsinki. All participants were fully informed of the study's purpose, content, and methodology. Participation was entirely voluntary, and the anonymity and confidentiality of participants were guaranteed. The questionnaire consisted of 25 items combining structured Likert-scale questions (1 = strongly disagree to 5 = strongly agree) and open-ended responses, organized into four thematic sections. These sections assessed perceived social benefits, such as ecological landscape enhancement, recreational value and cultural heritage preservation, economic impacts like job creation and tertiary industry development, environmental acceptance like technology preferences, tolerance toward construction disturbances, and causal relationships through scenario-based questions, such as shifts in support for site redevelopment. Variables were discretized into low, medium, and high categories, based on risk matrices and probability. Ethical considerations were rigorously addressed, and participants were provided with voluntary participation and data anonymization, according to established guidelines for social research. We collected data on the causal relationships between various factors, combined this with the actual situation to optimize it, clarified the causal relationships between various factors, adjusted the strength of the causal relationships, and reduced the complexity of the network, thus increasing the correlations between various factors and making the network structure more scientific. For parameter learning, the questionnaire data were first preprocessed and a two-dimensional risk matrix was established based on the severity of the impact that the risk factors may cause and the probability of risk occurrence. Reasoning analysis of the BNs was carried out using GeNie4.1, and the preprocessed data were used to perform parameter learning of the BN model to obtain the conditional probability distribution of each node variable (Fig. S1 ). 2.5 Uncertainty analysis The results were subjected to a complete uncertainty analysis including sensitivity analysis and Monte Carlo simulations. We used the Sobol index sensitivity method to analyze the sensitivity of the parameters. First-order and whole-order sensitivity coefficients were used to quantify the effects of the input variables on the output. The first-order sensitivity coefficient represents the contribution of a single parameter to the variation of output, while the whole-order sensitivity coefficient represents the sensitivity of a single parameter and its coupling with other parameters. The aim of these two coefficients is to separate the total variance of the objective function into the variance of single parameters and the variance of the interaction between multiple parameters 25 . We estimated the probability distributions of the input parameters and evaluated the uncertainty of the results using the Monte Carlo simulation 26 . Monte Carlo simulations use a priori known probability distributions of input variables to propagate the associated uncertainties through mathematical transformations to impact the derived quantities 27 . It is a stochastic simulation method that is used to numerically simulate the probability distribution of random variables. It estimates the output of complex systems using a large number of repeated random samples 28 . The Monte Carlo simulation carried out in Oracle Crystal Ball provides two types of results: the absolute uncertainty, in which the distribution of values within the 95% confidence interval (CI) for each type of impact is produced directly, based on 1,000 Monte Carlo simulations, and the level of confidence in the impact assessment of the two remediation alternatives. During the simulation, a comparison was repeated for each category and the data were selected randomly within the uncertainty range according to the uncertainty distribution 29 . In this study, we used the Monte Carlo simulation method to analyze the uncertainties of energy consumption, GHG emissions, the total emissions of three major air pollutants, including NO X , SO X , PM10, and HAPs emissions. Samples were selected randomly from the input dataset, assuming that all relevant parameters followed a normal distribution. 3 Results 3.1 Quantification of environmental impacts based on SEFA We quantified the energy consumption and GHG throughout the remediation process. A total of 71,944,800 million British thermal units (MMBTU) of energy was consumed during the restoration work, including the energy required for transportation and the operation of mechanical equipment, as well as the use of various materials throughout the construction phase. The carbon footprint and air pollutant emissions calculated by SEFA 3.0 encompassed the entire remediation lifecycle, including both on-site and off-site activities. The total GHG emissions were 10,163.06 tCO 2 -eq, predominantly associated with on-site activities, accounting for 86% of the total emissions. NO X emissions (429.07 t) were largely on-site, with 76% of the emissions occurring at the restoration site and only 24% off-site. In contrast, SO X emissions (129.89 t) were predominantly off-site, with only 2% occurring on-site and a substantial 97% off-site, and only 1% of emissions were from power grids. These data further emphasize the significance of off-site activities in the overall environmental impact, particularly in terms of SO X emissions. PM10 emissions (23.99 t) were 89% on-site and 11% off-site and HAPs emissions (0.85 t) were 27% on-site, 67% off-site, and 6% off-grid, requiring a multifaceted approach to pollution control (Fig. S2 ). The analysis of air pollutant emissions throughout the remediation process revealed significant variations in the release of SO X , NO X , PM10, and HAPs at different stages (Fig. 3 ). The NO X , SO X , PM10, and HAPs emissions were highest during soil remediation (427.88 t, 128.95 t, 23.95 t, and 0.83t, respectively). In the context of groundwater remediation, NO X and SO X emissions were relatively lower (216.23 kg and 63.68 kg, respectively), and PM10 emissions were even more marginal. The wastewater and solid waste treatment processes were associated with higher SO X and NO X emissions compared with PM10 (805.19 kg and 971.16 kg emitted, respectively). NO X , SO X , PM10, and HAPs emissions were lowest during the construction area-layout process (166.15 kg, 11.01 kg, 4.84 kg, and 2.3 kg, respectively). We compared the energy consumption and environmental footprints of TD and S/S. In this study, the main sources of TD’s environmental footprint were direct emissions during on-site restoration activities in scope 1 and other off-site emissions in scope 3, with environmental footprints generated by scope 3, scope 2, and scope 1 of 5.39–69.32%, 0–6.72%, and 23.96–94.61%, respectively (Fig. 4 A). The main source of the S/S desorption environmental footprint was other off-site emissions from scope 3, and a small amount of direct emissions from scope 1. The environmental footprints generated by scope 3 and scope 1 were 0.47–89.10% and 10.90–99.53%, respectively (Fig. 4 B). The energy consumption of TD was 51,808,820.45 MMBTU and the carbon emissions were 8,915.55 kgCO 2 -eq, while S/S consumed 555,706.31 MMBTU and its carbon emissions were 55,102.95 kgCO 2 -eq. Because of the difference in volume treated by the two remediation strategies, the environmental footprint of contaminated soil per unit of volume can be used as an important reference: TD consumed 1,260.42 MMBTU of energy per 1 m 3 of contaminated soil and its carbon emissions were 216.91 kgCO 2 -eq, while S/S consumed 360.03 MMBTU of energy and its carbon emissions were 35.72 kgCO 2 -eq. (A) TD; (B) S/S; (a) energy consumption; (b) GHG emissions; (c) air pollutants emissions (NO X , SO X , and PM10); and (d) HAPs emissions. 3.2 Socio-economic impact assessments of remediation alternatives based on MCA Regarding the social aspect, TD scores performed better in terms of installation and operational risks (SOC 1) and the robustness of sustainability assessment (SOC 5). In terms of the economic aspect, S/S scored lower against total economic cost (ECON 1), net present value (ECON 2), and operation time (ECON 5), but this was offset by employment opportunities and the net present value ratio. The remedial alternatives showed different results across various economic categories. TD cost 2,500.02 RMB per unit of contaminated soil, compared with 358.03 RMB for S/S. The complex operation, greater number of remediation links, and longer repair cycle also meant that TD provided more employment opportunities than S/S (Fig. 5 ). 3.3 Multi-criteria sustainability impact assessment based on BNs The BNs presented in Fig. 6 provide a detailed examination of the various benefits associated with the remediation of a contaminated oil refinery site and the subsequent establishment of a park on the former site. The social sustainability derived from the park development was rated as high with a probability of 71%, medium with a probability of 18%, and low with a probability of 10%. Park activities were highly beneficial, with a 71% probability of high benefit, followed by a 70% probability of high psychological function and a 71% probability of high service function. There was also a high probability (67%) of beneficial aesthetic function. Some aspects, however, such as traffic inconvenience and adverse effects, were more likely to be rated as having a high negative impact (30% and 56%, respectively). The economic benefits were rated as high with a probability 56%, medium with a probability of 27%, and low with a probability of 17%, indicating a strong positive impact on the economy. Employment opportunities also showed a high probability of being beneficial, with 73% rated as high. The development of the tertiary industries was similarly rated, with a 70% probability of high benefit. In terms of environmental benefits, aspects like overall satisfaction, plant communities, and coverage ratio were rated as highly beneficial, with probabilities of 88%, 77%, and 81%, respectively. The aesthetic function of the park also contributed positively to the environment, with an 85% probability of high benefit. These results highlighted the importance of the park development in enhancing the environmental quality and aesthetic appeal. 3.4 Uncertainty analysis of environmental impact The first-order and whole-order results of the different parameters were obtained based on the Sobol index sensitivity method, including the amount of material, the distance of material transportation, the operating hours of electric power equipment, and the operating hours of fuel equipment during the process of restoration. Among the total parameters, the sensitivity of material consumption was the highest, with a first-order sensitivity of 0.6049, indicating that material consumption was the key factor affecting GHG emissions. The first-order sensitivity of the material transport distance was 0.3387, indicating that material transport distance also influenced GHG emissions. In contrast, the first-order sensitivity of the hours of power equipment operation was 0.0498, indicating that this factor had a minimal contribution to GHG emissions. The number of operating hours of fuel equipment also had little effect on GHG emissions. Notably, the differences between the whole-order and first-order sensitivities for the four key parameters were very small, indicating that interactions between the parameters were almost non-existent and the influences of the parameters on the model output were significant but independent of the effects of other parameters (Fig. 7 ). We calculated the distributions of energy consumption and environmental impact values for each category based on 1,000 Monte Carlo simulations (Fig. 8 ). The energy consumption for the entire restoration project site was 71,944,800 MMBTU (standard deviation [SD] 7,194,694 MMBTU, 95% CI 60,111,933–83,778,667 MMBTU). The average GHG emissions resulting from the restoration project were 10,163 t (SD 1,016 t, 95% CI 8,491–11,835 t). The study also assessed the SO X , NO X , and PM10 emissions from the remediation project as 583 t (SD 0.58 t, 95% CI 470–697 t), and the mean HAPS emissions were 851 kg (SD 1.00 kg, 95% CI 687–1,016 kg). The coefficients of variation of the four variables were all < 10%, indicating that the significant variability and uncertainty were within an acceptable range. This result further supported the conclusion of low uncertainty in the carbon emission results in this study. 4 Discussion The present study reveals distinct environmental footprints for Thermal Desorption (TD) and Stabilization/Solidification (S/S). The primary impact of TD arises from on-site direct emissions, while S/S is dominated by off-site factors, including material transport and production. This distinction underscores that both on-site activities and supply chain logistics are critical concerns. To mitigate these impacts, strategic substitutions are key. For off-site emissions, optimizing transport by replacing long-distance trucking with rail can significantly reduce the carbon footprint 30 . For material-intensive methods like S/S, using low-carbon alternatives—such as substituting coal with biochar-based activated carbon—can reduce the overall environmental impact substantially and enhance long-term sustainability 31 . Interestingly, our study aligns with findings that lower energy use does not always guarantee lower GHG emissions in short-term remediation projects 29 . This highlights the need for comprehensive assessments that evaluate multiple indicators. Future remediation strategies must therefore adopt a holistic approach, prioritizing low-carbon materials (e.g., biochar composites, layered double hydroxides), optimizing transportation routes, and adopting renewable energy to minimize the cumulative environmental impact of large-scale projects. MCA scores in this study identified S/S as the more economical and safer option for workers, while TD offered greater employment opportunities. These results demonstrated that S/S-treated soils could maintain stability and effectiveness with long-term robustness, consistent with the findings of Wang et al. 32 . In contrast, despite its higher employment potential, TD carries significant risks of project failure due to its operational complexity and energy intensity 33 . Job creation during thermal remediation can provide important socio-economic benefits, as demonstrated in the post-industrial revitalization case studies analyzed by Cinelli et al. 34 , who showed that properly managed thermal remediation could catalyze local economic recovery when integrated with workforce development programs. The integration of social and economic indicators into remediation decision-making, as advocated by Gill et al. 35 , is essential for achieving holistic sustainability. Such integrated approaches would benefit from advanced decision-support systems that incorporate real-time monitoring data, as suggested by Xiao et al. 36 . Their work on adaptive remediation frameworks highlights how dynamic adjustments to treatment strategies based on ongoing performance data can significantly improve both environmental and economic outcomes. However, more studies are needed to evaluate the social sustainability dimensions of different remediation technologies, particularly in terms of their impacts on local communities and workforce development over extended time periods. Integrating remediation with site redevelopment is crucial for maximizing sustainability. Regenerating contaminated sites, particularly through preserving industrial heritage as seen in global case studies, offers significant environmental, social, and economic benefits over new construction by reducing carbon emissions and creating local value 37 . Public participation in China, however, presents unique challenges. Our findings suggest public perception is influenced more by information dissemination than direct involvement in decision-making, which contrasts with some Western literature emphasizing deep stakeholder collaboration 38 . This discrepancy highlights the need for culturally-sensitive engagement strategies and transparent communication to foster public trust. Future policies should prioritize meaningful community engagement to bridge the gap between technical solutions and societal expectations. We applied Bayesian Networks (BNs) to effectively manage the complex variables and uncertainty inherent in this process 22 . While useful, our analysis underscores a critical need for more interactive and accessible decision-support tools. Bridging the gap between technical analysis and public understanding is essential. To be truly sustainable, environmental management solutions must be socially robust and accepted by non-experts, not just technically sound 25 . 5 Conclusions This study evaluated the sustainability of remediation strategies for oil-contaminated land, focusing on TD and S/S. The results showed that GHG emissions mainly originate from on-site activities, but with a significant portion also occurring off-site. The quantity of materials was the most sensitive factor in terms of the environmental footprint, and there was no interaction among the different factors. An indicator set was developed to assess the sustainabilities of the remediation alternatives, and the results showed that S/S was more cost-effective and sustainable than TD, although TD provided more employment opportunities. This study emphasizes the importance of understanding the causal relationships in socio-economic systems and highlights the potential of BNs for future decision support. It also suggests examining structural uncertainties to improve the inclusion of expert knowledge in policy problem-solving. In summary, the results of this study support sustainable remediation practices and urban regeneration by providing a framework to evaluate the environmental, social, and economic impacts of transforming industrial sites into public spaces, advocating an approach to balance contamination reduction with minimal environmental impact. Declarations CRediT authorship contribution statement Yingluo Jia : Conceptualization, Methodology, Software, Writing – original draft. Xianglan Li : Validation, Writing – review & editing, Supervision, Funding acquisition. Hongzhen Zhang : Writing – review & editing, Funding acquisition, Resources, Data curation. Chunlong Zhang : Writing – review & editing. Yafei Wang : Data curation, Validation. Meijie Zhu : Data curation, Validation. Chunhui Sang : Data curation, Validation. Yuxin Nie : Data curation, Validation. Hao Meng : Data curation, Validation. Peng Liu : Data curation, Validation. Jingqi Dong : Data curation, Validation. All authors contributed critically to the draft and gave final approval for publication. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Acknowledgements This study was supported by The National Key Research and Development Program of China (Nos. 2022YFC3703300). References Kastanek, F. et al. Remediation of contaminated soils by thermal desorption; effect of benzoyl peroxide addition. J. Clean Prod. 125 , 309–313 (2016). Wei, K.-H. et al. Recent progress on in-situ chemical oxidation for the remediation of petroleum contaminated soil and groundwater. J. Hazard. Mater. 432 , 128738 (2022). Khan, M. B. et al. Potential sources and meteorological factors affecting PM2.5-bound polycyclic aromatic hydrocarbon levels in six main cities of northeastern Italy: an assessment of the related carcinogenic and mutagenic risks. Environ. Sci. Pollut. Res. 25 , 31987–32000 (2018). Zhao, C., Dong, Y., Feng, Y., Li, Y. & Dong, Y. Thermal desorption for remediation of contaminated soil: A review. Chemosphere 221 , 841–855 (2019). Wang, L. et al. Low-carbon and low-alkalinity stabilization/solidification of high-Pb contaminated soil. Chem. Eng. J. 351 , 418–427 (2018). Song, Y. et al. Environmental and socio-economic sustainability appraisal of contaminated land remediation strategies: A case study at a mega-site in China. Sci. Total Environ. 610 , 391–401 (2018). Cappuyns, V. Inclusion of social indicators in decision support tools for the selection of sustainable site remediation options. J. Environ. Manage. 184 , 45–56 (2016). Ellis, D. E. & Hadley, P. W. Sustainable remediation white paper-Integrating sustainable principles, practices, and metrics into remediation projects. Remediation (New York, N.Y.) 19 , 5–114 (2009). Khan, M. A. A., Qadir, Z., Asad, M., Kouzani, A. Z. & Parvez Mahmud, M. A. Environmental Footprint Assessment of a Cleanup at Hypothetical Contaminated Site. Appl. Sci.-Basel 11 , 4907 (2021). Beames, A., Broekx, S., Lookman, R., Touchant, K. & Seuntjens, P. Sustainability appraisal tools for soil and groundwater remediation: How is the choice of remediation alternative influenced by different sets of sustainability indicators and tool structures? Sci. Total Environ. 470 , 954–966 (2014). Huang, W.-Y. et al. Green and sustainable remediation (GSR) evaluation: framework, standards, and tool. A case study in Taiwan. Environ. Sci. Pollut. Res. 23 , 21712–21725 (2016). Hu, G. et al. Life cycle assessment of low-temperature thermal desorption-based technologies for drill cuttings treatment. J. Hazard. Mater. 401 , 123865 (2021). Hou, D., Gu, Q., Ma, F. & O’Connell, S. Life cycle assessment comparison of thermal desorption and stabilization/solidification of mercury contaminated soil on agricultural land. J. Clean Prod. 139 , 949–956 (2016). Harclerode, M. et al. Integrating the Social Dimension in Remediation Decision-Making: State of the Practice and Way Forward. Remediation Journal 26 , 11–42 (2015). Postle, M., Fenn, T., Grosso, A. & Steeds, J. Cost Benefit Analysis for Remediation of Land Contamination. Environment Agency (1999). Yang, D. China’s Environmental Law Fails to Effectively Ensure Fairness. in Chinese Research Perspectives on the Environment, Volume 1 vol. 1 (BRILL, United States, 2013). Petruzzi, N. M. A Case Study on the Evaluation and Implementation of Green and Sustainable Remediation Principles and Practices During a RCRA Corrective Action Cleanup. Ground Water Monit. Remediat. 31 , 63–71 (2011). Chen, H. Land use trade-offs associated with protected areas in China: Current state, existing evaluation methods, and future application of ecosystem service valuation. Sci. Total Environ. 711 , 134688 (2020). Said, Z. M. & Dindar, S. Key Challenges and Strategies in the Evaluation of Sustainable Urban Regeneration Projects: Insights from a Systematic Literature Review. Sustainability 16 , 9903 (2024). Rizzo, E. et al. Comparison of international approaches to sustainable remediation. J. Environ. Manage. 184 , 4–17 (2016). Fritsch, A. Elementary Bayesian Biostatistics. L. A. Moyé (2008). Boca Raton: Chapman & Hall/CRC. ISBN 978-1-58488-724-9. Biometrical Journal 51 , 546–547 (2009). Carriger, J. F. & Parker, R. A. Conceptual Bayesian networks for contaminated site ecological risk assessment and remediation support. J. Environ. Manage. 278 , 111478 (2021). Gao, W., Zeng, Z., Ma, X., Ke, Y. & Zhi, M. An improved hybrid structure learning strategy for Bayesian networks based on ensemble learning. Intelligent Data Analysis 27 , 1103–1120 (2023). Harris, M. J., Stinson, J. & Landis, W. G. A Bayesian Approach to Integrated Ecological and Human Health Risk Assessment for the South River, Virginia Mercury-Contaminated Site. Risk Analysis: An International Journal 37 , 1341–1357 (2017). Dekker, M. M. et al. Spread in climate policy scenarios unravelled. Nature 624 , (2023). Gregory, J. R., Noshadravan, A., Olivetti, E. A. & Kirchain, R. E. A Methodology for Robust Comparative Life Cycle Assessments Incorporating Uncertainty. Environ. Sci. Technol. 50 , 6397–6405 (2016). Jamroz, B. F. & Williams, D. F. Consistency in Monte Carlo Uncertainty Analyse. Metrologia 57 , (2020). Rajasekhar, B., Nambi, I. M. & Govindarajan, S. K. Human health risk assessment of ground water contaminated with petroleum PAHs using Monte Carlo simulations: A case study of an Indian metropolitan city. J. Environ. Manage. 205 , 183–191 (2018). Xiao, M., Li, X., Zhang, H., Meng, H. & Dong, J. Environmental impact assessment and remediation decision-making of a contaminated megasite: Combining LCA and IO-LCA. J. Clean Prod. 462 , 142586 (2024). Gallagher, P. M., Spatari, S. & Cucura, J. Hybrid life cycle assessment comparison of colloidal silica and cement grouted soil barrier remediation technologies. J. Hazard. Mater. 250 , 421–430 (2013). Wang, L. et al. Biochar composites: Emerging trends, field successes and sustainability implications. Soil Use Manage. 38 , 14–38 (2022). Wang, F., Wang, H. & Al-Tabbaa, A. Leachability and heavy metal speciation of 17-year old stabilised/solidified contaminated site soils. J. Hazard. Mater. 278 , 144–151 (2014). Paria, S. & Yuet, P. K. Solidification–stabilization of organic and inorganic contaminants using portland cement: a literature review. Environmental Reviews 14 , 217–255 (2006). Cinelli, M. et al. Supporting contaminated sites management with Multiple Criteria Decision Analysis: Demonstration of a regulation-consistent approach. J. Clean Prod. 316 , 128347 (2021). Gill, R. T., Thornton, S. F., Harbottle, M. J. & Smith, J. W. N. Sustainability assessment of electrokinetic bioremediation compared with alternative remediation options for a petroleum release site. J. Environ. Manage. 184 , 120–131 (2016). Xiao, M. et al. Qualitative and quantitative simulation of best management practices (BMPs) for contaminated megasite remediation using the SiteWiseTM tool. J. Environ. Manage. 360 , 121098 (2024). Hou, D. et al. Sustainable remediation and redevelopment of brownfield sites. Nat. Rev. Earth Environ. 4 , 271–286 (2023). Song, Y. et al. Environmental and socio-economic sustainability appraisal of contaminated land remediation strategies: A case study at a mega-site in China. Science of The Total Environment 610–611 , 391–401 (2018). Additional Declarations No competing interests reported. Supplementary Files BNUYingluoJiaetal.20250427.SI.pdf Cite Share Download PDF Status: Published Journal Publication published 26 Nov, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 30 Sep, 2025 Reviews received at journal 02 Aug, 2025 Reviews received at journal 27 Jul, 2025 Reviewers agreed at journal 23 Jul, 2025 Reviewers agreed at journal 23 Jul, 2025 Reviewers invited by journal 23 Jul, 2025 Editor assigned by journal 14 Jul, 2025 Editor invited by journal 30 Jun, 2025 Submission checks completed at journal 27 Jun, 2025 First submitted to journal 13 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6889334","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":489765199,"identity":"d37c0798-4d45-4f12-8442-db808b7ab06d","order_by":0,"name":"Yingluo Jia","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yingluo","middleName":"","lastName":"Jia","suffix":""},{"id":489765200,"identity":"9be57c47-d746-4e0a-8f0a-26fd4161eb4d","order_by":1,"name":"Xianglan Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAqElEQVRIiWNgGAWjYDACduaGAx8bwEwDIrUwMzYcnEmyFmZekrQYHGZsPGy7405iA3vzNgmGmjuEtUg2MzYczj3zLLGB51iZBMOxZ4S18DODtLQdTmyQyDGTALIJa2EDabEEaZF/Q6QWsC2MYFt4iNQC8svB3jOHjdt40ootEo4RocXgePPhDz93HJbtZz+88caHGiK0wAEbiEggQcMoGAWjYBSMAjwAAHp+OrlMN8NpAAAAAElFTkSuQmCC","orcid":"","institution":"Beijing Normal University","correspondingAuthor":true,"prefix":"","firstName":"Xianglan","middleName":"","lastName":"Li","suffix":""},{"id":489765202,"identity":"3a0e009c-3148-46a6-b8c2-7f9d7014abed","order_by":2,"name":"Hongzhen Zhang","email":"","orcid":"","institution":"Chinese Academy of Environmental Planning","correspondingAuthor":false,"prefix":"","firstName":"Hongzhen","middleName":"","lastName":"Zhang","suffix":""},{"id":489765203,"identity":"b45747a1-316c-4711-8aa2-dba96f139f74","order_by":3,"name":"Chunlong Zhang","email":"","orcid":"","institution":"University of Houston−Clear Lake","correspondingAuthor":false,"prefix":"","firstName":"Chunlong","middleName":"","lastName":"Zhang","suffix":""},{"id":489765204,"identity":"65c6123e-fb59-4137-924c-93eb18c433cd","order_by":4,"name":"Yafei Wang","email":"","orcid":"","institution":"Beijing Institute of Petrochemical Technology","correspondingAuthor":false,"prefix":"","firstName":"Yafei","middleName":"","lastName":"Wang","suffix":""},{"id":489765205,"identity":"05f06d15-cb7c-4289-99fb-dc032ac1d61e","order_by":5,"name":"Meijie Zhu","email":"","orcid":"","institution":"Hong Kong Chu Hai College","correspondingAuthor":false,"prefix":"","firstName":"Meijie","middleName":"","lastName":"Zhu","suffix":""},{"id":489765206,"identity":"4214dedb-9ff9-4c4b-92fd-9860d4847cec","order_by":6,"name":"Chunhui Sang","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Chunhui","middleName":"","lastName":"Sang","suffix":""},{"id":489765207,"identity":"7fc8e94f-6049-4d08-a1ff-1fc8ce56f709","order_by":7,"name":"Yuxin Nie","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yuxin","middleName":"","lastName":"Nie","suffix":""},{"id":489765208,"identity":"a6bb47c0-15c9-4822-9165-12d1b4564222","order_by":8,"name":"Hao Meng","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Meng","suffix":""},{"id":489765209,"identity":"f9cba596-b46c-4475-a0fe-a52e8ce15d3e","order_by":9,"name":"Peng Liu","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Liu","suffix":""},{"id":489765210,"identity":"d82712b1-fc91-4221-990e-4da615fa8a9c","order_by":10,"name":"Jingqi Dong","email":"","orcid":"","institution":"Chinese Academy of Environmental Planning","correspondingAuthor":false,"prefix":"","firstName":"Jingqi","middleName":"","lastName":"Dong","suffix":""}],"badges":[],"createdAt":"2025-06-13 15:08:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6889334/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6889334/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-25990-6","type":"published","date":"2025-11-26T15:57:43+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":87584140,"identity":"cbd3c9ba-0400-42d4-890d-b6ba83d6ca7d","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":213955,"visible":true,"origin":"","legend":"\u003cp\u003eSite map showing of remediation scope and sampling locations\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/2ee3d0587df27257d9226167.png"},{"id":87585346,"identity":"1afae7ad-7e38-43c0-94ed-6c778fbf4097","added_by":"auto","created_at":"2025-07-25 13:44:57","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":228299,"visible":true,"origin":"","legend":"\u003cp\u003eSystem boundaries for remediation alternatives: (a) TD alternatives; (b) S/S alternatives\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/0d9fe787dd590b58b35eddbe.png"},{"id":87584141,"identity":"ecfd1506-6ce8-4fe1-aa71-69864df1310b","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":85293,"visible":true,"origin":"","legend":"\u003cp\u003eNO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, PM10, and HAPs emissions during different remediation processes\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/ed6e0a27d589f2b6b9ee9e98.png"},{"id":87584470,"identity":"de821a79-def3-4580-aab7-ced344c57d38","added_by":"auto","created_at":"2025-07-25 13:36:57","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":114014,"visible":true,"origin":"","legend":"\u003cp\u003eSources of environmental footprints of contaminated soil remediation technologies\u003c/p\u003e\n\u003cp\u003e(A) TD; (B) S/S; (a) energy consumption; (b) GHG emissions; (c) air pollutants emissions (NO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, and PM10); and (d) HAPs emissions.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/feb125807da415f7791eac0a.png"},{"id":87584144,"identity":"d09b300c-9270-49a9-acb5-d3b1bac7858d","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":140059,"visible":true,"origin":"","legend":"\u003cp\u003eMCA scores of TD and S/S for economic and social indicators\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/f7280706a162a457376b2c0b.png"},{"id":87584150,"identity":"9265874d-37e0-44a2-b903-ac444f62fb78","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":338181,"visible":true,"origin":"","legend":"\u003cp\u003eBN parameter learning results of\u003cstrong\u003e \u003c/strong\u003ecomprehensive sustainability impact\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/9543f565f6679d9075089c01.png"},{"id":87584172,"identity":"334404a2-c288-4185-9d9a-376e8ef5ebf7","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":74963,"visible":true,"origin":"","legend":"\u003cp\u003eSobol index results for different variables\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/a22c2c634e3a47ee7f987444.png"},{"id":87584151,"identity":"57441f33-7a4e-4013-9f60-d065c972690a","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":143821,"visible":true,"origin":"","legend":"\u003cp\u003eUncertainty analysis based on 1,000 Monte Carlo simulations.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/8652e2ab2581b7f5bfcb2cbf.png"},{"id":97179335,"identity":"39412b6b-89e7-4db2-9057-497f72571f22","added_by":"auto","created_at":"2025-12-01 16:14:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2133386,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/0802b396-b7ff-4354-8321-311dcce55121.pdf"},{"id":87584147,"identity":"86c69e61-9b3a-4949-9418-7557745f3d4a","added_by":"auto","created_at":"2025-07-25 13:28:57","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":234333,"visible":true,"origin":"","legend":"","description":"","filename":"BNUYingluoJiaetal.20250427.SI.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6889334/v1/0583b63a1bc159b18a7a6fa0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Environmental and socio-economic sustainability assessment of remediation alternatives for a contaminated oil refinery site in Southern China","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eThe contamination of soils by oil and petroleum-based hydrocarbons as a result of accidental spills is causing increased environmental concern\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. In China, 77% of large-scale contaminated sites are polluted by organic contaminants including petroleum hydrocarbons\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. These pollutants have adverse impacts on human health and ecosystems, and the remediation of former refinery sites has attracted widespread attention\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Among the diverse remediation technologies, thermal desorption (TD) is noted for its high efficiency and broad applicability\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, while stabilization/solidification (S/S) is a cost-effective and robust method for immobilizing contaminants\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eRemediation professionals face many challenges in terms of incorporating sustainable remediation principles into real-world practice, especially in countries like China where the remediation industry is still in its infancy\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Selecting the optimal strategy requires a sustainable remediation framework that balances environmental, economic, and social aspects\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. While assessment frameworks exist, economic and social impacts are often underrepresented compared to environmental metrics\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Specialized tools like SEFA have advanced the quantification of environmental footprints, such as greenhouse gas emissions\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, and have been applied to petroleum-contaminated sites\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. However, there remains a lack of comprehensive, multi-criteria sustainability assessments for contaminated oil refinery sites in China, and existing appraisals can be subjective\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe present study selected a former refinery in Sichuan Province covering 52.32 hm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, associated with severe contamination of the soil and groundwater by petroleum hydrocarbons and heavy metals, necessitating efficient remediation to mitigate environmental and health risks. This study implemented TD and S/S as remediation strategies, and selected soil-remediation data samples for analysis. The objectives of this study were to: (1) conduct a specific and detailed quantitative environmental impact assessment of the remediation alternatives for a contaminated refinery site; (2) develop social and economic indicators and conduct a comparative socio-economic sustainability assessment of TD and S/S using MCA; and (3) conduct a probabilistic analysis of the causal relationship between public acceptance and reconstruction benefits by Bayesian networks (BNs).\u003c/p\u003e"},{"header":"2 Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1 Case study description\u003c/h2\u003e\n \u003cp\u003eThe refinery (total area approximately 52.32 hm\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) was located in Sichuan Province. It had been a significant environmental pollution concern to the locals and a demolition decision was made in 2013 to allow land redevelopment; a new park will be established on the site after appropriate remediation. The remediation scope is shown in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. The whole project needs to remediate 51,153.7 m\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e of soil. The main soils in the restoration area were miscellaneous fill soils, pebbles, and silt clay, and the permeability was poor. The thickness of the silt clay layer was 5.2\u0026ndash;13.3 m, with an average thickness of 8.4 m, and the maximum exposure depth was 9.1\u0026ndash;16.1 m, with an average depth of 13.42 m. According to the investigation results, the plot was polluted with aliphatic hydrocarbons (C\u003csub\u003e10\u003c/sub\u003e\u0026ndash;C\u003csub\u003e40\u003c/sub\u003e), benzo[a]pyrene, benzo[a]anthracene, dibenz[a,h]anthracene, and the heavy metals cobalt, lead, and arsenic. Based on the formation conditions, the characteristics of the pollutants and the spatial distribution of the remediation area, TD and S/S restoration models were adopted.\u003c/p\u003e\n \u003cp\u003eThe environmental footprint-accounting boundary included the entire process of TD and S/S from the beginning of remediation to the end, including construction-site layout, soil remediation, groundwater remediation, and wastewater disposal (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). The environmental footprint was generated by the use and transportation of fuel oil, electrical machinery, and construction materials and chemicals.\u003c/p\u003e\n \u003cp\u003eSEFA 3.0 (EPA, USA) is a GSR framework decision-support tool that quantifies energy use, GHG emissions, and air pollutants across remediation phases to identify high-impact activities. It also provides qualitative descriptions of the impacts of remediation activities on the land and ecosystems\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Additional activities, such as field experiments and MCA, can be conducted to supplement the assessment results. SEFA integrates energy, air emissions, water, and materials using lifecycle inventory data to model site-specific scenarios. In this study, TD and S/S strategies were evaluated during the remediation process using SEFA 3.0, including analyzing inputs such as electricity, fuel, and materials to generate outputs for energy consumption, GHG emissions, and air pollutants, including nitrogen oxide (NO\u003csub\u003eX\u003c/sub\u003e), sulfur oxide (SO\u003csub\u003eX\u003c/sub\u003e), particulate matter with a diameter\u0026thinsp;\u0026le;\u0026thinsp;10 \u0026micro;m (PM10), and hazardous air pollutants (HAPs).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2 Remediation strategies\u003c/h2\u003e\n \u003cp\u003eTwo major site-remediation technologies, TD and S/S, were used. TD is a process that heats the contaminated soil to volatilize pollutants and then collects them for treatment, ensuring effective remediation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. The equipment produces high-temperature flue gas through gas combustion on the outer wall of the TD chamber, transfers the heat to the material in the chamber, and achieves soil heating\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e up to 800℃, at which almost all the soil can be vaporized and separated. The high-temperature exhaust gas from TD is treated successively and then introduced into the combustion area of the burner for disposal. S/S removes the contaminated soil and fixes the contaminant by adding a stabilizer to reduce the risk of migration\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. TD has no need to excavate and is thus a flexible process with good mobility and can be reused. S/S employs agents such as silicate cement to encapsulate heavy metals, and once the curing process is complete and the concentrations of heavy metals in both acid and water leachates are shown to be below the stipulated standards, the treated material can then be disposed of through \u003cem\u003ein situ\u003c/em\u003e or barrier containment and landfilling. The environmental footprints of the two remediation technologies were calculated separately and the sources of the environmental footprints were divided into three scopes: scope 1 involves the direct emission of on-site restoration activities; scope 2 involves the electricity generation; and scope 3 involves other off-site emissions generated from the production, disposal, and transportation of upstream and downstream products and wastes.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e2.3 Social and economic sustainability assessment\u003c/h2\u003e\n \u003cp\u003eTo assess the sustainability of the two major site-remediation technologies, we selected a set of indicators to assess the economic and social sustainabilities. The indicator categories included economic costs and benefits, worker health and safety, communities and community involvement, and public acceptance. Economic indicators were represented by ECON 1\u0026ndash;ECON 5 and Social indicators were represented by SOC 1\u0026ndash;SOC 5 (Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e). All the indicators were derived from existing guidelines and realistic concerns about the context of the remediation industry in China. The different options were assigned a ranked score of 1\u0026ndash;5 framed against an idealized scenario, with a lower score indicating a more-sustainable option and a higher score indicating a less-sustainable option.\u003c/p\u003e\n \u003cp\u003eMulti-criteria models can help remediation practitioners to evaluate multiple conflicting criteria in order to make sustainable remediation decisions about contaminated sites\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. MCA is a method that is increasingly used for remediation sustainability assessments, to support decision-making. In this study, we used a similar approach to that described by Postle et al.\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Both quantitative and qualitative indicators were used to inform the MCA (Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e), and qualitative indicators were assigned a score of 1\u0026ndash;5, with lower scores indicating superior outcomes:\u003c/p\u003e\n \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n \u003cdiv class=\"EquationNumber\"\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1753449748.png\" width=\"651\" height=\"91\"\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere \u0026nbsp;is MCA score, \u0026nbsp;is the input value for each indicator, and is the priority weighting for each indicator. The indicators were normalized by weighing the input values against the maximum values across all treatment options and multiplied to provide a score of 0\u0026ndash;100. The priority weighting system was based on the concerns of stakeholders.\u003c/p\u003e\n \u003cp\u003eSocial impacts were given emphasis in the sustainability assessment. Worker safety was identified as a fundamental social indicator, in light of its regulatory enforcement in many countries, where occupational health standards are increasingly stringent\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Public acceptance, measured through community involvement and satisfaction, reflects stakeholder engagement; however, community participation in China is often limited, unless required by environmental impact assessments\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Equality was an optional indicator to address concerns of environmental injustice, which are particularly relevant in China where disparities in pollution exposure and remediation benefits have been documented\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Local community impacts, such as dust, noise, and traffic congestion, were also considered, with environmental supervision ensuring compliance with national standards, as well as increased traffic volume associated with off-site impacts, including harmful air emissions, noise pollution, and road degradation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eEconomic impacts were evaluated by balancing costs and benefits. Direct benefits included increased land value, while indirect benefits encompassed employment and local business opportunities\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. The economic benefits included not only financial expenditures, but also temporal and technological risks. Shorter remediation timelines are often prioritized to accelerate redevelopment and maximize economic returns, reflecting the rapid pace of urban regeneration\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. To enhance its applicability, the indicator set was structured into core and optional elements, allowing flexibility to adapt to diverse project contexts and stakeholder priorities\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e2.4 Overall sustainability appraisal\u003c/h2\u003e\n \u003cp\u003eA BN is a graphical probabilistic model used to represent and manipulate uncertain knowledge, and is employed to examine interactions and probability dependencies between variables\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. It consists of a set of random variables and their conditional dependencies, which are usually represented by a directed acyclic graph. BNs are often constructed to reflect causality and follow a directed path from cause to effect\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The goal of BN structure learning is to construct a network using expert knowledge or observational datasets, so that the learned network can maximize the expression of the complex associations between random variables\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. BNs include parent and child nodes, with each node representing a random variable: parent nodes represent input parameters and child nodes receive inputs from one or more parent nodes, while the directed edges indicate conditional dependencies between the variables. The graphical engine and concept mapping style creates a transparent causal model that can be evaluated with data and expert knowledge at all phases of the model-building process\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eBNs incorporate the deterministic and stochastic aspects of complex systems, explicitly consider uncertainty in the model inputs, and provide probabilistic predictions with measures of the importance of the input variables (sensitivity analysis). Input parameters are represented as probability distributions, which are derived directly from monitoring data. Parent, child, and endpoint nodes are discretized into ranked states, which allows the evaluation of the combined effects of multiple stressors, including categorical factors or factors with varying units of measurement\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eIn this study, BNs provided a framework to assess citizens\u0026rsquo; expectations and opinions about the remediation and transformation of the oil refinery. A structured questionnaire was developed to investigate public expectations and perceptions regarding the remediation and redevelopment of the contaminated oil refinery site. A total of 40 participants were recruited, targeting residents of diverse backgrounds, including local residents, employees of nearby industries, environmental professionals, and students, to capture societal perspectives. The questionnaires were distributed via community events, email, and social media platforms such as WeChat, to ensure broad demographic coverage and minimize selection bias. The study protocol was approved by the Faculty of Geographical Science, Beijing Normal University. In accordance with the regulations of Beijing Normal University, this study is classified as a routine assessment project and, therefore, does not require approval from an Ethics Committee or Institutional Review Board. The study does not involve animal or human clinical trials and is not unethical. The research was conducted in line with the ethical principles outlined in the Declaration of Helsinki. All participants were fully informed of the study\u0026apos;s purpose, content, and methodology. Participation was entirely voluntary, and the anonymity and confidentiality of participants were guaranteed.\u003c/p\u003e\n \u003cp\u003eThe questionnaire consisted of 25 items combining structured Likert-scale questions (1\u0026thinsp;=\u0026thinsp;strongly disagree to 5\u0026thinsp;=\u0026thinsp;strongly agree) and open-ended responses, organized into four thematic sections. These sections assessed perceived social benefits, such as ecological landscape enhancement, recreational value and cultural heritage preservation, economic impacts like job creation and tertiary industry development, environmental acceptance like technology preferences, tolerance toward construction disturbances, and causal relationships through scenario-based questions, such as shifts in support for site redevelopment. Variables were discretized into low, medium, and high categories, based on risk matrices and probability. Ethical considerations were rigorously addressed, and participants were provided with voluntary participation and data anonymization, according to established guidelines for social research.\u003c/p\u003e\n \u003cp\u003eWe collected data on the causal relationships between various factors, combined this with the actual situation to optimize it, clarified the causal relationships between various factors, adjusted the strength of the causal relationships, and reduced the complexity of the network, thus increasing the correlations between various factors and making the network structure more scientific. For parameter learning, the questionnaire data were first preprocessed and a two-dimensional risk matrix was established based on the severity of the impact that the risk factors may cause and the probability of risk occurrence. Reasoning analysis of the BNs was carried out using GeNie4.1, and the preprocessed data were used to perform parameter learning of the BN model to obtain the conditional probability distribution of each node variable (Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e2.5 Uncertainty analysis\u003c/h2\u003e\n \u003cp\u003eThe results were subjected to a complete uncertainty analysis including sensitivity analysis and Monte Carlo simulations. We used the Sobol index sensitivity method to analyze the sensitivity of the parameters. First-order and whole-order sensitivity coefficients were used to quantify the effects of the input variables on the output. The first-order sensitivity coefficient represents the contribution of a single parameter to the variation of output, while the whole-order sensitivity coefficient represents the sensitivity of a single parameter and its coupling with other parameters. The aim of these two coefficients is to separate the total variance of the objective function into the variance of single parameters and the variance of the interaction between multiple parameters\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eWe estimated the probability distributions of the input parameters and evaluated the uncertainty of the results using the Monte Carlo simulation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Monte Carlo simulations use \u003cem\u003ea priori\u003c/em\u003e known probability distributions of input variables to propagate the associated uncertainties through mathematical transformations to impact the derived quantities\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. It is a stochastic simulation method that is used to numerically simulate the probability distribution of random variables. It estimates the output of complex systems using a large number of repeated random samples\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The Monte Carlo simulation carried out in Oracle Crystal Ball provides two types of results: the absolute uncertainty, in which the distribution of values within the 95% confidence interval (CI) for each type of impact is produced directly, based on 1,000 Monte Carlo simulations, and the level of confidence in the impact assessment of the two remediation alternatives. During the simulation, a comparison was repeated for each category and the data were selected randomly within the uncertainty range according to the uncertainty distribution\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In this study, we used the Monte Carlo simulation method to analyze the uncertainties of energy consumption, GHG emissions, the total emissions of three major air pollutants, including NO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, PM10, and HAPs emissions. Samples were selected randomly from the input dataset, assuming that all relevant parameters followed a normal distribution.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Quantification of environmental impacts based on SEFA\u003c/h2\u003e\u003cp\u003eWe quantified the energy consumption and GHG throughout the remediation process. A total of 71,944,800\u0026nbsp;million British thermal units (MMBTU) of energy was consumed during the restoration work, including the energy required for transportation and the operation of mechanical equipment, as well as the use of various materials throughout the construction phase. The carbon footprint and air pollutant emissions calculated by SEFA 3.0 encompassed the entire remediation lifecycle, including both on-site and off-site activities. The total GHG emissions were 10,163.06 tCO\u003csub\u003e2\u003c/sub\u003e-eq, predominantly associated with on-site activities, accounting for 86% of the total emissions.\u003c/p\u003e\u003cp\u003eNO\u003csub\u003eX\u003c/sub\u003e emissions (429.07 t) were largely on-site, with 76% of the emissions occurring at the restoration site and only 24% off-site. In contrast, SO\u003csub\u003eX\u003c/sub\u003e emissions (129.89 t) were predominantly off-site, with only 2% occurring on-site and a substantial 97% off-site, and only 1% of emissions were from power grids. These data further emphasize the significance of off-site activities in the overall environmental impact, particularly in terms of SO\u003csub\u003eX\u003c/sub\u003e emissions. PM10 emissions (23.99 t) were 89% on-site and 11% off-site and HAPs emissions (0.85 t) were 27% on-site, 67% off-site, and 6% off-grid, requiring a multifaceted approach to pollution control (Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe analysis of air pollutant emissions throughout the remediation process revealed significant variations in the release of SO\u003csub\u003eX\u003c/sub\u003e, NO\u003csub\u003eX\u003c/sub\u003e, PM10, and HAPs at different stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The NO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, PM10, and HAPs emissions were highest during soil remediation (427.88 t, 128.95 t, 23.95 t, and 0.83t, respectively). In the context of groundwater remediation, NO\u003csub\u003eX\u003c/sub\u003e and SO\u003csub\u003eX\u003c/sub\u003e emissions were relatively lower (216.23 kg and 63.68 kg, respectively), and PM10 emissions were even more marginal. The wastewater and solid waste treatment processes were associated with higher SO\u003csub\u003eX\u003c/sub\u003e and NO\u003csub\u003eX\u003c/sub\u003e emissions compared with PM10 (805.19 kg and 971.16 kg emitted, respectively). NO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, PM10, and HAPs emissions were lowest during the construction area-layout process (166.15 kg, 11.01 kg, 4.84 kg, and 2.3 kg, respectively).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe compared the energy consumption and environmental footprints of TD and S/S. In this study, the main sources of TD\u0026rsquo;s environmental footprint were direct emissions during on-site restoration activities in scope 1 and other off-site emissions in scope 3, with environmental footprints generated by scope 3, scope 2, and scope 1 of 5.39\u0026ndash;69.32%, 0\u0026ndash;6.72%, and 23.96\u0026ndash;94.61%, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). The main source of the S/S desorption environmental footprint was other off-site emissions from scope 3, and a small amount of direct emissions from scope 1. The environmental footprints generated by scope 3 and scope 1 were 0.47\u0026ndash;89.10% and 10.90\u0026ndash;99.53%, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003eThe energy consumption of TD was 51,808,820.45 MMBTU and the carbon emissions were 8,915.55 kgCO\u003csub\u003e2\u003c/sub\u003e-eq, while S/S consumed 555,706.31 MMBTU and its carbon emissions were 55,102.95 kgCO\u003csub\u003e2\u003c/sub\u003e-eq.\u0026nbsp;Because of the difference in volume treated by the two remediation strategies, the environmental footprint of contaminated soil per unit of volume can be used as an important reference: TD consumed 1,260.42 MMBTU of energy per 1 m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e of contaminated soil and its carbon emissions were 216.91 kgCO\u003csub\u003e2\u003c/sub\u003e-eq, while S/S consumed 360.03 MMBTU of energy and its carbon emissions were 35.72 kgCO\u003csub\u003e2\u003c/sub\u003e-eq.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e(A) TD; (B) S/S; (a) energy consumption; (b) GHG emissions; (c) air pollutants emissions (NO\u003csub\u003eX\u003c/sub\u003e, SO\u003csub\u003eX\u003c/sub\u003e, and PM10); and (d) HAPs emissions.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Socio-economic impact assessments of remediation alternatives based on MCA\u003c/h2\u003e\u003cp\u003eRegarding the social aspect, TD scores performed better in terms of installation and operational risks (SOC 1) and the robustness of sustainability assessment (SOC 5). In terms of the economic aspect, S/S scored lower against total economic cost (ECON 1), net present value (ECON 2), and operation time (ECON 5), but this was offset by employment opportunities and the net present value ratio. The remedial alternatives showed different results across various economic categories. TD cost 2,500.02 RMB per unit of contaminated soil, compared with 358.03 RMB for S/S. The complex operation, greater number of remediation links, and longer repair cycle also meant that TD provided more employment opportunities than S/S (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Multi-criteria sustainability impact assessment based on BNs\u003c/h2\u003e\u003cp\u003eThe BNs presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e provide a detailed examination of the various benefits associated with the remediation of a contaminated oil refinery site and the subsequent establishment of a park on the former site. The social sustainability derived from the park development was rated as high with a probability of 71%, medium with a probability of 18%, and low with a probability of 10%. Park activities were highly beneficial, with a 71% probability of high benefit, followed by a 70% probability of high psychological function and a 71% probability of high service function. There was also a high probability (67%) of beneficial aesthetic function. Some aspects, however, such as traffic inconvenience and adverse effects, were more likely to be rated as having a high negative impact (30% and 56%, respectively). The economic benefits were rated as high with a probability 56%, medium with a probability of 27%, and low with a probability of 17%, indicating a strong positive impact on the economy. Employment opportunities also showed a high probability of being beneficial, with 73% rated as high. The development of the tertiary industries was similarly rated, with a 70% probability of high benefit. In terms of environmental benefits, aspects like overall satisfaction, plant communities, and coverage ratio were rated as highly beneficial, with probabilities of 88%, 77%, and 81%, respectively. The aesthetic function of the park also contributed positively to the environment, with an 85% probability of high benefit. These results highlighted the importance of the park development in enhancing the environmental quality and aesthetic appeal.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.4 Uncertainty analysis of environmental impact\u003c/h2\u003e\u003cp\u003eThe first-order and whole-order results of the different parameters were obtained based on the Sobol index sensitivity method, including the amount of material, the distance of material transportation, the operating hours of electric power equipment, and the operating hours of fuel equipment during the process of restoration. Among the total parameters, the sensitivity of material consumption was the highest, with a first-order sensitivity of 0.6049, indicating that material consumption was the key factor affecting GHG emissions. The first-order sensitivity of the material transport distance was 0.3387, indicating that material transport distance also influenced GHG emissions. In contrast, the first-order sensitivity of the hours of power equipment operation was 0.0498, indicating that this factor had a minimal contribution to GHG emissions. The number of operating hours of fuel equipment also had little effect on GHG emissions. Notably, the differences between the whole-order and first-order sensitivities for the four key parameters were very small, indicating that interactions between the parameters were almost non-existent and the influences of the parameters on the model output were significant but independent of the effects of other parameters (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe calculated the distributions of energy consumption and environmental impact values for each category based on 1,000 Monte Carlo simulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The energy consumption for the entire restoration project site was 71,944,800 MMBTU (standard deviation [SD] 7,194,694 MMBTU, 95% CI 60,111,933\u0026ndash;83,778,667 MMBTU). The average GHG emissions resulting from the restoration project were 10,163 t (SD 1,016 t, 95% CI 8,491\u0026ndash;11,835 t). The study also assessed the SO\u003csub\u003eX\u003c/sub\u003e, NO\u003csub\u003eX\u003c/sub\u003e, and PM10 emissions from the remediation project as 583 t (SD 0.58 t, 95% CI 470\u0026ndash;697 t), and the mean HAPS emissions were 851 kg (SD 1.00 kg, 95% CI 687\u0026ndash;1,016 kg). The coefficients of variation of the four variables were all \u0026lt;\u0026thinsp;10%, indicating that the significant variability and uncertainty were within an acceptable range. This result further supported the conclusion of low uncertainty in the carbon emission results in this study.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eThe present study reveals distinct environmental footprints for Thermal Desorption (TD) and Stabilization/Solidification (S/S). The primary impact of TD arises from on-site direct emissions, while S/S is dominated by off-site factors, including material transport and production. This distinction underscores that both on-site activities and supply chain logistics are critical concerns.\u003c/p\u003e\u003cp\u003eTo mitigate these impacts, strategic substitutions are key. For off-site emissions, optimizing transport by replacing long-distance trucking with rail can significantly reduce the carbon footprint\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. For material-intensive methods like S/S, using low-carbon alternatives\u0026mdash;such as substituting coal with biochar-based activated carbon\u0026mdash;can reduce the overall environmental impact substantially and enhance long-term sustainability\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eInterestingly, our study aligns with findings that lower energy use does not always guarantee lower GHG emissions in short-term remediation projects\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. This highlights the need for comprehensive assessments that evaluate multiple indicators. Future remediation strategies must therefore adopt a holistic approach, prioritizing low-carbon materials (e.g., biochar composites, layered double hydroxides), optimizing transportation routes, and adopting renewable energy to minimize the cumulative environmental impact of large-scale projects.\u003c/p\u003e\u003cp\u003eMCA scores in this study identified S/S as the more economical and safer option for workers, while TD offered greater employment opportunities. These results demonstrated that S/S-treated soils could maintain stability and effectiveness with long-term robustness, consistent with the findings of Wang et al.\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In contrast, despite its higher employment potential, TD carries significant risks of project failure due to its operational complexity and energy intensity\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Job creation during thermal remediation can provide important socio-economic benefits, as demonstrated in the post-industrial revitalization case studies analyzed by Cinelli et al.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, who showed that properly managed thermal remediation could catalyze local economic recovery when integrated with workforce development programs.\u003c/p\u003e\u003cp\u003eThe integration of social and economic indicators into remediation decision-making, as advocated by Gill et al.\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, is essential for achieving holistic sustainability. Such integrated approaches would benefit from advanced decision-support systems that incorporate real-time monitoring data, as suggested by Xiao et al.\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Their work on adaptive remediation frameworks highlights how dynamic adjustments to treatment strategies based on ongoing performance data can significantly improve both environmental and economic outcomes. However, more studies are needed to evaluate the social sustainability dimensions of different remediation technologies, particularly in terms of their impacts on local communities and workforce development over extended time periods.\u003c/p\u003e\u003cp\u003eIntegrating remediation with site redevelopment is crucial for maximizing sustainability. Regenerating contaminated sites, particularly through preserving industrial heritage as seen in global case studies, offers significant environmental, social, and economic benefits over new construction by reducing carbon emissions and creating local value\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003ePublic participation in China, however, presents unique challenges. Our findings suggest public perception is influenced more by information dissemination than direct involvement in decision-making, which contrasts with some Western literature emphasizing deep stakeholder collaboration\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. This discrepancy highlights the need for culturally-sensitive engagement strategies and transparent communication to foster public trust. Future policies should prioritize meaningful community engagement to bridge the gap between technical solutions and societal expectations.\u003c/p\u003e\u003cp\u003eWe applied Bayesian Networks (BNs) to effectively manage the complex variables and uncertainty inherent in this process\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. While useful, our analysis underscores a critical need for more interactive and accessible decision-support tools. Bridging the gap between technical analysis and public understanding is essential. To be truly sustainable, environmental management solutions must be socially robust and accepted by non-experts, not just technically sound\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e"},{"header":"5 Conclusions","content":"\u003cp\u003eThis study evaluated the sustainability of remediation strategies for oil-contaminated land, focusing on TD and S/S. The results showed that GHG emissions mainly originate from on-site activities, but with a significant portion also occurring off-site. The quantity of materials was the most sensitive factor in terms of the environmental footprint, and there was no interaction among the different factors. An indicator set was developed to assess the sustainabilities of the remediation alternatives, and the results showed that S/S was more cost-effective and sustainable than TD, although TD provided more employment opportunities. This study emphasizes the importance of understanding the causal relationships in socio-economic systems and highlights the potential of BNs for future decision support. It also suggests examining structural uncertainties to improve the inclusion of expert knowledge in policy problem-solving.\u003c/p\u003e\u003cp\u003eIn summary, the results of this study support sustainable remediation practices and urban regeneration by providing a framework to evaluate the environmental, social, and economic impacts of transforming industrial sites into public spaces, advocating an approach to balance contamination reduction with minimal environmental impact.\u003c/p\u003e\u003cp\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCRediT authorship contribution statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eYingluo Jia\u003c/strong\u003e: Conceptualization, Methodology, Software, Writing \u0026ndash; original draft. \u003cstrong\u003eXianglan Li\u003c/strong\u003e: Validation, Writing \u0026ndash; review \u0026amp; editing, Supervision, Funding acquisition. \u003cstrong\u003eHongzhen Zhang\u003c/strong\u003e: Writing \u0026ndash; review \u0026amp; editing, Funding acquisition, Resources, Data curation. \u003cstrong\u003eChunlong Zhang\u003c/strong\u003e: Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eYafei Wang\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003eMeijie Zhu\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003eChunhui Sang\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003eYuxin Nie\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003eHao Meng\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003ePeng Liu\u003c/strong\u003e: Data curation, Validation. \u003cstrong\u003eJingqi Dong\u003c/strong\u003e: Data curation, Validation. All authors contributed critically to the draft and gave final approval for publication.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of competing interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analysed during the current study available from the corresponding author on reasonable request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by The National Key Research and Development Program of China (Nos. 2022YFC3703300).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eKastanek, F. \u003cem\u003eet al.\u003c/em\u003e Remediation of contaminated soils by thermal desorption; effect of benzoyl peroxide addition. \u003cem\u003eJ. Clean Prod.\u003c/em\u003e \u003cstrong\u003e125\u003c/strong\u003e, 309\u0026ndash;313 (2016).\u003c/li\u003e\n\u003cli\u003eWei, K.-H. \u003cem\u003eet al.\u003c/em\u003e Recent progress on in-situ chemical oxidation for the remediation of petroleum contaminated soil and groundwater. \u003cem\u003eJ. Hazard. Mater.\u003c/em\u003e \u003cstrong\u003e432\u003c/strong\u003e, 128738 (2022).\u003c/li\u003e\n\u003cli\u003eKhan, M. B. \u003cem\u003eet al.\u003c/em\u003e Potential sources and meteorological factors affecting PM2.5-bound polycyclic aromatic hydrocarbon levels in six main cities of northeastern Italy: an assessment of the related carcinogenic and mutagenic risks. \u003cem\u003eEnviron. Sci. Pollut. Res.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 31987\u0026ndash;32000 (2018).\u003c/li\u003e\n\u003cli\u003eZhao, C., Dong, Y., Feng, Y., Li, Y. \u0026amp; Dong, Y. Thermal desorption for remediation of contaminated soil: A review. \u003cem\u003eChemosphere\u003c/em\u003e \u003cstrong\u003e221\u003c/strong\u003e, 841\u0026ndash;855 (2019).\u003c/li\u003e\n\u003cli\u003eWang, L. \u003cem\u003eet al.\u003c/em\u003e Low-carbon and low-alkalinity stabilization/solidification of high-Pb contaminated soil. \u003cem\u003eChem. Eng. J.\u003c/em\u003e \u003cstrong\u003e351\u003c/strong\u003e, 418\u0026ndash;427 (2018).\u003c/li\u003e\n\u003cli\u003eSong, Y. \u003cem\u003eet al.\u003c/em\u003e Environmental and socio-economic sustainability appraisal of contaminated land remediation strategies: A case study at a mega-site in China. \u003cem\u003eSci. Total Environ.\u003c/em\u003e \u003cstrong\u003e610\u003c/strong\u003e, 391\u0026ndash;401 (2018).\u003c/li\u003e\n\u003cli\u003eCappuyns, V. Inclusion of social indicators in decision support tools for the selection of sustainable site remediation options. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e184\u003c/strong\u003e, 45\u0026ndash;56 (2016).\u003c/li\u003e\n\u003cli\u003eEllis, D. E. \u0026amp; Hadley, P. W. Sustainable remediation white paper-Integrating sustainable principles, practices, and metrics into remediation projects. \u003cem\u003eRemediation (New York, N.Y.)\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 5\u0026ndash;114 (2009).\u003c/li\u003e\n\u003cli\u003eKhan, M. A. A., Qadir, Z., Asad, M., Kouzani, A. Z. \u0026amp; Parvez Mahmud, M. A. Environmental Footprint Assessment of a Cleanup at Hypothetical Contaminated Site. \u003cem\u003eAppl. Sci.-Basel\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 4907 (2021).\u003c/li\u003e\n\u003cli\u003eBeames, A., Broekx, S., Lookman, R., Touchant, K. \u0026amp; Seuntjens, P. Sustainability appraisal tools for soil and groundwater remediation: How is the choice of remediation alternative influenced by different sets of sustainability indicators and tool structures? \u003cem\u003eSci. Total Environ.\u003c/em\u003e \u003cstrong\u003e470\u003c/strong\u003e, 954\u0026ndash;966 (2014).\u003c/li\u003e\n\u003cli\u003eHuang, W.-Y. \u003cem\u003eet al.\u003c/em\u003e Green and sustainable remediation (GSR) evaluation: framework, standards, and tool. A case study in Taiwan. \u003cem\u003eEnviron. Sci. Pollut. Res.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 21712\u0026ndash;21725 (2016).\u003c/li\u003e\n\u003cli\u003eHu, G. \u003cem\u003eet al.\u003c/em\u003e Life cycle assessment of low-temperature thermal desorption-based technologies for drill cuttings treatment. \u003cem\u003eJ. Hazard. Mater.\u003c/em\u003e \u003cstrong\u003e401\u003c/strong\u003e, 123865 (2021).\u003c/li\u003e\n\u003cli\u003eHou, D., Gu, Q., Ma, F. \u0026amp; O\u0026rsquo;Connell, S. Life cycle assessment comparison of thermal desorption and stabilization/solidification of mercury contaminated soil on agricultural land. \u003cem\u003eJ. Clean Prod.\u003c/em\u003e \u003cstrong\u003e139\u003c/strong\u003e, 949\u0026ndash;956 (2016).\u003c/li\u003e\n\u003cli\u003eHarclerode, M. \u003cem\u003eet al.\u003c/em\u003e Integrating the Social Dimension in Remediation Decision-Making: State of the Practice and Way Forward. \u003cem\u003eRemediation Journal\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 11\u0026ndash;42 (2015).\u003c/li\u003e\n\u003cli\u003ePostle, M., Fenn, T., Grosso, A. \u0026amp; Steeds, J. Cost Benefit Analysis for Remediation of Land Contamination. \u003cem\u003eEnvironment Agency\u003c/em\u003e (1999).\u003c/li\u003e\n\u003cli\u003eYang, D. China\u0026rsquo;s Environmental Law Fails to Effectively Ensure Fairness. in \u003cem\u003eChinese Research Perspectives on the Environment, Volume 1\u003c/em\u003e vol. 1 (BRILL, United States, 2013).\u003c/li\u003e\n\u003cli\u003ePetruzzi, N. M. A Case Study on the Evaluation and Implementation of Green and Sustainable Remediation Principles and Practices During a RCRA Corrective Action Cleanup. \u003cem\u003eGround Water Monit. Remediat.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 63\u0026ndash;71 (2011).\u003c/li\u003e\n\u003cli\u003eChen, H. Land use trade-offs associated with protected areas in China: Current state, existing evaluation methods, and future application of ecosystem service valuation. \u003cem\u003eSci. Total Environ.\u003c/em\u003e \u003cstrong\u003e711\u003c/strong\u003e, 134688 (2020).\u003c/li\u003e\n\u003cli\u003eSaid, Z. M. \u0026amp; Dindar, S. Key Challenges and Strategies in the Evaluation of Sustainable Urban Regeneration Projects: Insights from a Systematic Literature Review. \u003cem\u003eSustainability\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 9903 (2024).\u003c/li\u003e\n\u003cli\u003eRizzo, E. \u003cem\u003eet al.\u003c/em\u003e Comparison of international approaches to sustainable remediation. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e184\u003c/strong\u003e, 4\u0026ndash;17 (2016).\u003c/li\u003e\n\u003cli\u003eFritsch, A. Elementary Bayesian Biostatistics. L. A. Moy\u0026eacute; (2008). Boca Raton: Chapman \u0026amp; Hall/CRC. ISBN 978-1-58488-724-9. \u003cem\u003eBiometrical Journal\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 546\u0026ndash;547 (2009).\u003c/li\u003e\n\u003cli\u003eCarriger, J. F. \u0026amp; Parker, R. A. Conceptual Bayesian networks for contaminated site ecological risk assessment and remediation support. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e278\u003c/strong\u003e, 111478 (2021).\u003c/li\u003e\n\u003cli\u003eGao, W., Zeng, Z., Ma, X., Ke, Y. \u0026amp; Zhi, M. An improved hybrid structure learning strategy for Bayesian networks based on ensemble learning. \u003cem\u003eIntelligent Data Analysis\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1103\u0026ndash;1120 (2023).\u003c/li\u003e\n\u003cli\u003eHarris, M. J., Stinson, J. \u0026amp; Landis, W. G. A Bayesian Approach to Integrated Ecological and Human Health Risk Assessment for the South River, Virginia Mercury-Contaminated Site. \u003cem\u003eRisk Analysis: An International Journal\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 1341\u0026ndash;1357 (2017).\u003c/li\u003e\n\u003cli\u003eDekker, M. M. \u003cem\u003eet al.\u003c/em\u003e Spread in climate policy scenarios unravelled. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e624\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eGregory, J. R., Noshadravan, A., Olivetti, E. A. \u0026amp; Kirchain, R. E. A Methodology for Robust Comparative Life Cycle Assessments Incorporating Uncertainty. \u003cem\u003eEnviron. Sci. Technol.\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 6397\u0026ndash;6405 (2016).\u003c/li\u003e\n\u003cli\u003eJamroz, B. F. \u0026amp; Williams, D. F. Consistency in Monte Carlo Uncertainty Analyse. \u003cem\u003eMetrologia\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eRajasekhar, B., Nambi, I. M. \u0026amp; Govindarajan, S. K. Human health risk assessment of ground water contaminated with petroleum PAHs using Monte Carlo simulations: A case study of an Indian metropolitan city. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e205\u003c/strong\u003e, 183\u0026ndash;191 (2018).\u003c/li\u003e\n\u003cli\u003eXiao, M., Li, X., Zhang, H., Meng, H. \u0026amp; Dong, J. Environmental impact assessment and remediation decision-making of a contaminated megasite: Combining LCA and IO-LCA. \u003cem\u003eJ. Clean Prod.\u003c/em\u003e \u003cstrong\u003e462\u003c/strong\u003e, 142586 (2024).\u003c/li\u003e\n\u003cli\u003eGallagher, P. M., Spatari, S. \u0026amp; Cucura, J. Hybrid life cycle assessment comparison of colloidal silica and cement grouted soil barrier remediation technologies. \u003cem\u003eJ. Hazard. Mater.\u003c/em\u003e \u003cstrong\u003e250\u003c/strong\u003e, 421\u0026ndash;430 (2013).\u003c/li\u003e\n\u003cli\u003eWang, L. \u003cem\u003eet al.\u003c/em\u003e Biochar composites: Emerging trends, field successes and sustainability implications. \u003cem\u003eSoil Use Manage.\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 14\u0026ndash;38 (2022).\u003c/li\u003e\n\u003cli\u003eWang, F., Wang, H. \u0026amp; Al-Tabbaa, A. Leachability and heavy metal speciation of 17-year old stabilised/solidified contaminated site soils. \u003cem\u003eJ. Hazard. Mater.\u003c/em\u003e \u003cstrong\u003e278\u003c/strong\u003e, 144\u0026ndash;151 (2014).\u003c/li\u003e\n\u003cli\u003eParia, S. \u0026amp; Yuet, P. K. Solidification\u0026ndash;stabilization of organic and inorganic contaminants using portland cement: a literature review. \u003cem\u003eEnvironmental Reviews\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 217\u0026ndash;255 (2006).\u003c/li\u003e\n\u003cli\u003eCinelli, M. \u003cem\u003eet al.\u003c/em\u003e Supporting contaminated sites management with Multiple Criteria Decision Analysis: Demonstration of a regulation-consistent approach. \u003cem\u003eJ. Clean Prod.\u003c/em\u003e \u003cstrong\u003e316\u003c/strong\u003e, 128347 (2021).\u003c/li\u003e\n\u003cli\u003eGill, R. T., Thornton, S. F., Harbottle, M. J. \u0026amp; Smith, J. W. N. Sustainability assessment of electrokinetic bioremediation compared with alternative remediation options for a petroleum release site. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e184\u003c/strong\u003e, 120\u0026ndash;131 (2016).\u003c/li\u003e\n\u003cli\u003eXiao, M. \u003cem\u003eet al.\u003c/em\u003e Qualitative and quantitative simulation of best management practices (BMPs) for contaminated megasite remediation using the SiteWiseTM tool. \u003cem\u003eJ. Environ. Manage.\u003c/em\u003e \u003cstrong\u003e360\u003c/strong\u003e, 121098 (2024).\u003c/li\u003e\n\u003cli\u003eHou, D. \u003cem\u003eet al.\u003c/em\u003e Sustainable remediation and redevelopment of brownfield sites. \u003cem\u003eNat. Rev. Earth Environ.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 271\u0026ndash;286 (2023).\u003c/li\u003e\n\u003cli\u003eSong, Y. \u003cem\u003eet al.\u003c/em\u003e Environmental and socio-economic sustainability appraisal of contaminated land remediation strategies: A case study at a mega-site in China. \u003cem\u003eScience of The Total Environment\u003c/em\u003e \u003cstrong\u003e610\u0026ndash;611\u003c/strong\u003e, 391\u0026ndash;401 (2018).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Contaminated site remediation, Multi-criteria assessment, Decision-making, Oil refinery site","lastPublishedDoi":"10.21203/rs.3.rs-6889334/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6889334/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRefineries are an important industrial legacy, emitting large amounts of petroleum hydrocarbons and highlighting the need for the green and sustainable remediation of contaminated oil refinery sites globally. This study aimed to evaluate the environmental and socio-economic sustainability of thermal desorption (TD) and stabilization/solidification (S/S) strategies for remediating a contaminated oil refinery site in Sichuan Province, China. We calculated the energy consumption, greenhouse gas emissions, and air pollutants across different remediation phases, and quantified the sustainability of the remediation using multi criteria analysis (MCA) and Bayesian networks (BNs). TD consumed 518,088,020.45\u0026nbsp;million British thermal units (MMBTU) of energy and emitted 8,915.55 kgCO\u003csub\u003e2\u003c/sub\u003e-eq carbon, compared with 555,706.31 MMBTU and 5,102.95 kgCO\u003csub\u003e2\u003c/sub\u003e-eq, respectively, for S/S. Socio-economic appraisal showed that S/S was associated with lower economic costs, better worker safety, and greater sustainability, while TD provided more employment opportunities. BN analysis further predicted a 71% probability of high social benefits and a 56% probability of economic benefits from converting the site into a public park. These results highlight the need for integrated strategies balancing environmental remediation, economic viability, and community engagement, to provide a framework for the sustainable urban regeneration of industrial legacies.\u003c/p\u003e","manuscriptTitle":"Environmental and socio-economic sustainability assessment of remediation alternatives for a contaminated oil refinery site in Southern China","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-25 13:28:52","doi":"10.21203/rs.3.rs-6889334/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-30T15:17:18+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-02T11:10:18+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-27T16:17:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"229904992127637452002695947003368689340","date":"2025-07-23T12:43:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"41886742676901748096518043047573469107","date":"2025-07-23T05:58:16+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-23T05:46:38+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-14T11:55:59+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-07-01T02:00:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-28T03:34:16+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-06-13T14:55:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d3fb8d58-5d38-485a-9ab6-32b5f7a76675","owner":[],"postedDate":"July 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":51997972,"name":"Earth and environmental sciences/Environmental sciences"},{"id":51997973,"name":"Earth and environmental sciences/Environmental social sciences"}],"tags":[],"updatedAt":"2025-12-01T16:09:16+00:00","versionOfRecord":{"articleIdentity":"rs-6889334","link":"https://doi.org/10.1038/s41598-025-25990-6","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-11-26 15:57:43","publishedOnDateReadable":"November 26th, 2025"},"versionCreatedAt":"2025-07-25 13:28:52","video":"","vorDoi":"10.1038/s41598-025-25990-6","vorDoiUrl":"https://doi.org/10.1038/s41598-025-25990-6","workflowStages":[]},"version":"v1","identity":"rs-6889334","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6889334","identity":"rs-6889334","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}