LCA is a bottom-up, ISO-standardized methodology that systematically quantifies environmental emissions, resource consumption, and associated impacts across the life cycle of a product or service, ranging from raw material extraction (cradle) to end-of-life disposal or recovery (grave) [Figure 2]^[20-22]. In environmental remediation, LCA enables a rigorous, detailed comparison of alternative remediation technologies by evaluating their cumulative energy demand, global warming potential, and other environmental impact categories^[14]. LCA comprises five interrelated phases, namely goal and scope definition, life cycle inventory analysis, life cycle impact assessment, interpretation, and reporting. This framework promotes a comprehensive and transparent analysis of environmental performance^[23]. However, LCA is data-intensive and often requires extensive primary data collection supplemented by high-quality secondary data^[24,25]. The results can be sensitive to methodological choices, including system boundaries, allocation procedures, and impact assessment models, thereby necessitating careful interpretation and scenario analysis^[26,27].

I-O analysis is a top-down macroeconomic approach that quantifies environmental emissions and resource flows across interconnected economic sectors. Its widely adopted hybrid form in environmental assessment is environmentally extended I-O LCA (EIO-LCA), which estimates emissions based on intersectoral monetary transactions^[19]. A three‑tier accounting framework is typically employed, encompassing direct emissions, energy‑related indirect emissions, and broader supply-chain emissions. This structure makes I-O analysis particularly effective for evaluating environmental impacts at the economy‑wide or sectoral scales^[28]. However, its dependence on aggregated national or regional economic datasets limits the level of detail and technological specificity that can be achieved when assessing individual projects, products, or processes. Consequently, I-O analysis has limited applicability in project-specific decision-making contexts^[29].

The IPCC calculation method provides a standardized and globally recognized framework for GHG emission estimations. Total emissions are calculated by multiplying activity data such as the volume of consumed fuel or the quantity of treated waste by the corresponding emission factors. Developed under the auspices of the IPCC, this method ensures consistency and comparability in national GHG inventories and corporate reporting^[19]. Its tiered structure offers flexibility, where default emissions factors can be replaced with technology-specific, geographically localized, or higher-tier data to improve estimation accuracy^[30]. However, the reliability of the resulting estimates strongly depends on the precision, representativeness, and quality of the selected activity data and emission factors. The use of generic or outdated coefficients can introduce considerable uncertainty into carbon-footprint estimates^[19].

HLCA integrates process-level, detail-oriented data from conventional LCA with the economy-wide coverage of I-O analysis. By combining the strengths of these two approaches, HLCA aims to balance technological specificity and system completeness. It reduces the truncation errors often arising from process-based LCA due to system boundary cut-offs and offers higher technological resolution than purely I-O-based models^[31,32]. These characteristics make HLCA particularly well-suited for evaluating complex, multicomponent systems such as integrated remediation strategies, where significant indirect or supply-chain contributions may otherwise be overlooked^[33-35]. Despite these advantages, several methodological challenges remain. These include consistently and accurately defining hybrid integration boundaries between processes and I-O components, reconciling data disparities such as temporal, geographical, and technological resolution, and avoiding the double counting of emissions. Careful methodological harmonization and transparent reporting are required to address these issues^[34].

At the project level, carbon accounting serves two primary purposes. First, it disaggregates remediation project into constituent processes to identify emission hotspots for targeted mitigation measures. Second, it enables the comparison of alternative remediation strategies, supporting the selection of options with greater environmental co-benefits^[17]. Several assessment methodologies and decision-support tools have been proposed for support these purposes. Prominent examples include the Spreadsheet for Environmental Footprint Analysis from the USEPA, the LCA framework endorsed by the Sustainable Remediation Forum, and specialized carbon-accounting calculators such as the soil remediation tool, SiteWise, and the Tauw CO₂ Calculator^[13,36]. These tools estimate emissions (CO₂eq) by applying emission factors to activity data such as fuel and electricity consumption.

Real-world case studies showed that these tools can support low-carbon remediation decision-making. For example, a comparative assessment of bioremediation and chemical oxidation at a petroleum-hydrocarbon contaminated site found that bioremediation generated only 15.64 tons of CO₂eq compared with 118.4 tons produced via chemical oxidation. In addition to lower emissions, bioremediation improved soil fertility and biological activity, indicating broader sustainability benefits^[37]. Another study reported that integrating renewable energy sources into remediation measures reduced the overall environmental footprint by over 68%, highlighting the significant potential of operational decarbonization strategies^[38].

Among these methods, LCA is the most comprehensive and widely used framework because it evaluates environmental impacts across all stages of remediation projects^[14]. LCA follows a standardized five-phase framework. Site-scale LCA studies have consistently shown that in situ remediation techniques typically have considerably lower carbon footprint than conventional excavation and off-site disposal approaches^[18,36,39]. LCA can also identify high-impact activities within remediation systems. For example, long-distance transportation of treatment residues, particularly via shipping, is major contributor to project-level carbon footprint^[40,41]. Similarly, thermal desorption of mercury-contaminated soil generates 357 kg of CO₂eq per ton of soil, providing a useful benchmark for performance evaluation^[42]. LCA also supports process optimization. Studies have shown that small adjustments to standard practices can considerably lower carbon footprint. For example, adding biochar into solidification or stabilization mixtures or operating thermal desorption systems at lower treatment temperatures can reduce emission by more than half. Such analyses provide valuable guidance for fine-tuning remediation processes and minimizing climate impact^[43,44]. However, the reliability of LCA heavily depends on data quality and methodological choices^[45]. Inconsistent system boundaries, outdated databases, and unrealistic assumptions can yield inaccurate results. To address these challenges, a standardized nine-step LCA protocol has been proposed for remediation projects. This protocol includes goal and scope definition, functional unit definition, system boundary establishment, project metrics selection, project inventory compilation, impact assessment, sensitivity and uncertainty analysis, interpretation of inventory analysis and impact-assessment results, and reporting. This protocol emphasizes transparent data sourcing, consistent functional-unit selection, and thorough sensitivity analysis to enhance the reliability and comparability of studies^[14]. Simplified carbon calculators are ideal for rapid screening but often overlook embodied carbon emissions associated with materials and equipment. To address these issues, more advanced life cycle-based tools have been developed. The application of such a tool in a large-scale project comparing soil washing with excavation and landfill showed that soil washing reduced the carbon footprint by 14%. The analysis also identified specific emission mitigation strategies such as local material sourcing and improved fuel efficiency during earthworks^[46].

Site-scale carbon-footprint assessments are supported by a growing body of empirical case studies and validated tools; however, scaling up carbon accounting to regional and macro scales is relatively limited. Contrary to project-level analyses, few remediation-specific, data-driven analyses have quantified carbon footprint at such broader spatial scales. Consequently, existing studies rely largely on methodological frameworks, methodological extrapolations, and forward-looking analytical perspectives than on mature evidence bases. IO-LCA is often suggested as a suitable method for regional-scale carbon accounting. This hybrid method combines process-based LCA data with economic I-O tables that capture interdependencies across economic sectors^[17]. Its key theoretical advantage is its ability to define comprehensive system boundaries, which can capture upstream supply-chain emissions that may be truncated in process-based LCA^[13,47]. However, the application of IO-LCA to contaminated-site remediation remains scarce. A seminal study of brownfield redevelopment in San Francisco employed IO-LCA to estimate the carbon footprint of remediation activities across the city’s sites. The study concluded that brownfield redevelopment combined with sustainable remediation practices could reduce net GHG emissions by 519 t CO₂eq over a 70-year period, equivalent to 14% of the city’s emissions in 2010. The results demonstrated clear climate benefits of redeveloping contaminated land relative to pristine greenfield sites^[15].

In China, regional-scale carbon-footprint studies related to remediation are even more limited. One inventory of 48 remediation projects in Beijing cataloged the prevalence of in situ and ex situ technologies, providing a foundational dataset that could support future regional carbon-accounting measures. In addition, concepts such as City Carbon Maps and Carbon Networks (CCN) have been proposed as visualization tools for tracking material, energy, and emission flows among contamination sources, remediation sites, and waste disposal facilities^[48,49]. However, these concepts remain conceptual and have not been empirically validated via remediation-specific applications. Regional-scale assessments can help identify low-carbon remediation technologies specific to a region. Regions with similar contamination profiles, geological conditions, and socioeconomic settings can benefit from tailored remediation approaches. For instance, bioremediation is preferred for widespread oil contamination in the Niger Delta due to its lower estimated environmental footprint and cost; however, rigorous carbon accounting has to be performed to validate these assumptions^[50].

A potential framework for regional carbon accounting involves a multitiered approach. First, standardized emission factors are developed for commonly employed remediation technologies using LCA. Second, a regional inventory of contamination sites is established and remediation pathways are assigned based on-site characteristics. Finally, section-wide emissions are estimated by combining site inventories, soil volumes, technology-specific emission factors, and other relevant parameters. Such a framework can provide a scalable and scientifically grounded model for quantifying the contribution of remediation activities to regional carbon budgets.

At the national and global scales, dedicated carbon-footprint assessments for the remediation sector are currently lacking. However, methodologies from environmental economics such as multiregional I-O (MRIO) modeling have been used to quantify emissions embodied in international trade and national consumption^[51-53]. Applying these macro-level tools to contaminated-site remediation can enable top-down estimates of sector-wide emissions and support the integration of remediation activities into net-zero strategies^[17].

While these methods provide the analytical backbone for carbon-footprint assessment, their application to contaminated-site remediation introduces several methodological challenges that considerably influence the accuracy of and comparability across studies. Current challenges include inconsistent reporting of emission scopes (with Scope 3 emissions often partially or fully overlooked); inadequate accounting for embodied carbon associated with remediation amendments, reagents, and equipment manufacture; and substantial variability in the handling of transportation activities and off-site disposal processes. Moreover, avoided emissions and substitution benefits are either excluded or credited inconsistently, whereas temporal boundaries for long-term processes are rarely standardized, affecting technology comparisons. Most studies lack uncertainty and sensitivities analyses, hindering robust interpretation and cross-study comparability. To improve methodological consistency, the transparent reporting of life cycle stages, explicit justification of exclusions, sensitivity testing, and separate presentation of gross and net emissions will have to be undertaken.

The most notable distinction of remediation-related carbon emissions is the choice between in situ (treatment in place) and ex situ (excavation and off-site treatment) strategies. Ex situ technologies, particularly dig-and-dump, have consistently ranked among the most carbon-intensive remediation approaches^[58-60]. A substantial carbon burden is created by their heavy reliance on fossil fuels for large-scale earthmoving, long-distance transportation of contaminated materials, and long-term landfill management (including associated methane emissions). In contrast, disruptive logistics are avoided by in situ technologies, resulting in a lower baseline carbon footprint and supporting the transition toward sustainable remediation^[1].

The inherent characteristics of a contaminated site establish the fundamental boundary conditions that strongly influence technology selection and the resulting carbon footprint^[86]. Contaminant type, concentration, and distribution are primary drivers. For example, dense, deep zones of chlorinated solvent contamination may require energy-intensive in situ thermal treatment, whereas shallow, biodegradable petroleum hydrocarbon contamination may be suitable for low-carbon enhanced aerobic bioremediation^[87,88]. Site geology and hydrogeology are equally important. Low-permeability clay layers can considerably increase the energy required for pump-and-treat systems or the quantity of reagents needed for in situ chemical oxidation to achieve effective contact, thereby increasing the carbon footprint^[89,90]. In contrast, a permeable and homogeneous aquifer facilitates more efficient treatment. The scale and depth of contamination directly affect material and energy demands, with large and deep plumes inevitably requiring more resources than small and shallow plumes^[87,89]. Finally, site location and accessibility influence transportation distances for equipment, materials, excavated soil, and waste, which can considerably affect emissions. Off-site disposal or liquid reagent delivery at remote sites can be disproportionately carbon-intensive^[91].

Decisions made during design and implementation provide crucial opportunities for carbon mitigation. Treatment duration is a key factor, particularly for technologies with continuous energy demands. Substantial carbon emissions can be accumulated by a poorly optimized pump-and-treat system operating for decades, even if its hourly energy consumption is modest^[80]. Therefore, remediation strategies that achieve cleanup objectives more rapidly, despite higher initial emissions, may result in a lower overall lifecycle carbon footprint. This highlights the importance of system optimization and adaptive management. The use of real-time monitoring data to adjust pumping rates, reagent injection, or heating parameters dynamically can reduce substantial energy and material waste. Technology configuration and operational design are also important factors. For example, electrical resistive heating used for in situ thermal treatment in regions with low-carbon electricity grids may have a different carbon footprint than gas-fired thermal conduction heating^[92]. Similarly, the choice of reagents and amendments such as different oxidants or locally sourced versus commercially produced carbon substrates for bioremediation can considerably influence emissions^[93].

Remediation projects operate within broader economic and infrastructural systems that influence their indirect emissions. The carbon intensity of the local electricity grid is a vital external factor affecting electricity-dependent technologies such as pump-and-treat systems. The same electrical load can have markedly different carbon footprint depending on whether electricity is generated from renewable sources or fossil fuels^[94]. This factor can alter the comparative carbon ranking of remediation technologies across regions. Supply-chain emissions associated with manufacturing and transport of remediation materials, including steel sheet piles, ZVI, activated carbon, and geosynthetics, contribute considerably to embodied carbon. Although these emissions are often overlooked, they are captured in comprehensive LCAs^[95]. In addition, current regulatory frameworks and waste management practices can influence carbon footprint^[96]. Regulations that impose overly conservative cleanup standards or prohibit on-site treatment can favor higher carbon pathways such as excavation and off-site disposal^[97,98]. In contrast, policies that promote the beneficial reuse of treated soils or provide access to renewable energy directly enable low-carbon remediation solutions.

In summary, although technology selection establishes the initial carbon signature of a remediation project, the final carbon footprint depends on a complex interplay of site characteristics, design efficiency, and external conditions [Figure 5]. Among these, the carbon intensity of the energy supply is the most influential factor. The source of electricity (fossil fuel-based grid power versus low-carbon renewable alternatives) strongly influences the carbon footprint of energy-intensive processes such as pump-and-treat and thermal desorption^[17,99]. Transportation logistics are also a major contributor because emissions scale directly with distances between remediation sites, material suppliers, and waste disposal facilities. Therefore, local sourcing of materials and using nearby disposal facilities can help reduce carbon footprint. Project duration further influences the emission profile. Technologies characterized by short periods of high-intensity energy consumption have different temporal carbon distribution than those requiring lower but sustained energy inputs over long operating periods^[100]. In addition, the embodied carbon in chemical reagents, binders, and soil amendments represents a significant upstream emission source that must be integrated into comprehensive LCAs^[101]. Effective carbon management in remediation thus requires a system-based perspective that evaluates and optimizes across all of these interacting factors.

A primary decarbonization strategy is the transition from conventional, often extensive treatment methods toward smarter, more efficient remediation systems. Over-treatment (e.g., excessive use of energy, chemicals, or operational effort) represents a significant yet frequently overlooked source of carbon emissions. Precision remediation addresses this by employing real-time monitoring, machine learning, and adaptive management to dynamically tailor remediation activities to actual contaminant distribution and concentration. For example, energy inputs in thermal remediation systems can be adjusted based on real-time subsurface temperature and contaminant data rather than being operated continuously at high power^[102,103]. Similarly, bioremediation can be optimized by monitoring geochemical parameters and supplying nutrients or electron donors only when microbial activity declines^[1,63]. This “just enough, just-in-time” approach minimizes energy and resource consumption, directly reducing the carbon footprint without compromising remediation objectives. Although pilot studies have demonstrated energy savings of 20%-40%, independent validation across diverse site conditions remains limited^[63]. However, high costs for sensor networks and the need for specialized expertise constrain broader adoption. Aggressive optimization may extend project duration or risk incomplete treatment if safety margins are reduced; however, these trade-offs are rarely quantified in existing research.

The carbon intensity of the local electricity grid is a major determinant of remediation-related emissions. Therefore, integrating on-site renewable energy offers a direct pathway for decarbonization. Solar photovoltaic arrays are increasingly deployed on remediated brownfields or containment caps to power remediation activities^[104,105]. This approach reduces dependence on grid electricity and lower transmission-related losses. For larger, long-term projects, hybrid systems incorporating solar, wind, or geothermal energy are being explored^[106]. Repurposing remediation sites as temporary renewable-energy farms (e.g., solar power) can create revenue streams that can support cleanup activities^[1]. This creates a virtuous cycle, simultaneously aligning environmental and economic benefits. However, the extent of carbon reduction strongly depends on local grid carbon intensity and system intermittency, which often requires battery storage or grid backup, and thus adds cost and embodied carbon. For short-duration projects, the carbon payback period associated with manufacturing renewable-energy equipment may exceed the operational benefits achieved. Using remediated land for solar farms may delay or preclude alternative redevelopment options (e.g., housing), representing a land-use trade-off that is seldom discussed in remediation-focused studies^[107].

The production and transportation of remediation amendments and chemicals are carbon-intensive processes. An emerging strategy is to replace these inputs with sustainable alternatives. Biochar, a carbon-rich material produced via biomass pyrolysis, is a premier example^[108]. It serves a dual function: it sequesters carbon in soil over extended periods, acting as a direct carbon sink, whereas its high surface area and porosity enable effective immobilization of heavy metals and organic contaminants. This approach reflects circular-economy principles by converting agricultural or forestry waste into valuable remediation materials^[1]. However, biochar quality varies considerably with feedstock characteristics and pyrolysis conditions, and the lack of standardization hinders regulatory acceptance and performance predictability. Similarly, the use of waste-derived products such as treated industrial slags or green-synthesized nanoparticles reduces dependence on virgin resources^[109]. The development of green chemicals such as life-cycle-assessed chemical oxidants or biodegradable surfactants further minimizes the embodied carbon and environmental toxicity of remediation additives^[110,111].

The mobilization of heavy equipment and transportation of materials or contaminated soil account for a substantial portion of a project’s direct emissions, particularly in ex situ remediation. Green logistics strategies aim to reduce these impacts via route optimization for soil hauling to reduce fuel consumption, scheduling equipment to avoid idling, and adopting lower-emission vehicles such as those powered by biodiesel or electricity^[1,112]. The shift toward in situ remediation also inherently eliminates most transportation-related emissions by treating contamination directly at the site. When excavation is unavoidable, on-site treatment or local beneficial reuse of treated soil can considerably reduce transportation requirements; this considerably reduces GHG emissions and local air pollutants. However, biodiesel availability and performance under cold weather conditions may be limited in some regions, and electric heavy equipment for earthmoving remains commercially immature. On-site treatment may require additional land area and permitting, which can be challenging at space-constrained urban sites.

The widespread adoption of low-carbon remediation strategies requires their formal integration into remedy-selection processes^[17,21]. This can be achieved using policy instruments and sophisticated decision-support tools^[113]. LCA is being increasingly required or encouraged to provide quantitative assessments of the environmental impacts of different remediation alternatives, including global warming potential^[114]. These data can be incorporated into multicriteria decision analysis (MCDA) frameworks. Traditionally, MCDA frameworks have balanced criteria such as cost, remediation efficiency, and implementation time. Recently, carbon footprint and sustainability metrics have been incorporated as additional decision criteria^[115]. This structured approach enables transparent and defensible decision-making by explicitly integrating decarbonization objectives into traditional remediation planning, making them technically feasible and environmentally superior. Despite conceptual progress, mandatory carbon accounting remains rare in actual remedy selection. LCA requires data and expertise that are often unavailable to site managers, whereas the weighting of MCDA criteria remains inherently subjective because stakeholders may assign different levels of importance to carbon reduction, cost, or remediation speed. Regulatory inertia and liability concerns, particularly concerns that a low-carbon remedy will fail to meet cleanup standards, are significant barriers. Furthermore, prioritizing carbon reduction may favor slower‑acting remedies such as monitored natural attenuation, which may conflict with the need for rapid risk reduction. These trade‑offs must be transparently addressed within decision-making frameworks.

Individual strategies can achieve measurable carbon reductions; however, their combined application may generate synergistic benefits that exceed the sum of their individual contributions. For example, precision remediation coupled with on‑site solar-power scheduling may further reduce carbon footprint compared with solar energy alone. Similarly, biochar‑based stabilization combined with phytoremediation creates a closed‑loop, carbon‑negative system that promotes net carbon sequestration. Integrating green logistics with low‑carbon amendments may yield substantial emission reductions while generating additional synergistic benefits^[116,117]. However, not all couplings are beneficial. For example, combining high‑intensity precision control with renewable energy for short‑term projects may increase equipment emissions without delivering proportional benefits. Thus, the effectiveness of coupling decarbonization strategies should be evaluated using project‑specific LCA rather than assuming additive benefits.