Techno-economic and life cycle assessment of an integrated electrocoagulation process for sustainable treatment of arsenic and fluoride contaminated groundwater

Abstract

The present study evaluated the environmental and economic sustainability of an electrocoagulation-based system for arsenic and fluoride removal through life cycle analysis (LCA) and techno-economic assessment (TEA). The impact of different treatment capacities of the EC process has been evaluated. With the increase in scaling up the EC system from lab-scale (1.7 L) to large-scale (650 L), a reduction in environmental footprints across all ReCiPe midpoint categories (e.g. global warming potential (GWP) 5.38 to 2.63 kg CO₂ eq. (51%)) has been observed. Further, endpoint analysis indicated significant damage to the ecosystem (9.3 × 10⁻⁹ species per year) and resources ($0116). Interestingly, the negative endpoint human health impact values (−7.3 × 10⁻⁶ disability adjusted life years (DALY) for large-scale operation) suggest potential health benefits from treated water. Sensitivity and Monte Carlo analyses confirmed the robustness and reliability of the results, while utilizing carbon free electricity sources further reduced the impacts (GWP 36.7% less for solar). TEA analysis confirms the profitability of the EC treatment process since the net present value (NPV), internal rate of return (IRR), payback period (PB) and profitability index (PI) of the best scenario (3 shift operation with sludge utilization) are as INR 450.7 Lakhs ($0.51 million), 49.3%, 2.17 years and 1.85, respectively. Utilizing solar energy increases the capital expenditure (CAPEX) of the process by 64.2% but reduces the operational expenditure (OPEX) by 11.4%. Despite the higher initial investment for the use of solar energy, the overall scenario remains economically profitable. Overall, the integration of LCA and TEA highlights the feasibility of scaling up the electrocoagulation process as a sustainable and cost-effective solution for real-world applications.

---

Water impact

This study demonstrates the environmental and economic feasibility of scaling up an electrocoagulation-based integrated treatment system for arsenic- and fluoride-contaminated groundwater. It aligns with the Sustainable Development Goals (SDG-6, SDG-13, and SDG-3). Future implementation, especially with renewable energy integration, can drive sustainable and impactful water treatment for community application.

1. Introduction

Sustainable treatment of arsenic and fluoride containing groundwater is a crucial challenge across the globe. More than 220 million are at risk of diseases caused by arsenic contamination while more than 200 million are exposed to the fluoride contaminated groundwater.^1,2 Electrocoagulation (EC) based processes have shown efficient simultaneous removal of arsenic and fluoride.³ EC-based treatment has been reported as an environmentally friendly and low-cost method for the treatment of arsenic and fluoride-containing groundwater compared to aluminium hydroxide/oxide-based nanoparticle-based adsorptive treatment.⁴ However, most of the EC-based studies have shown efficient performance over the lab-scale small reactor set-ups only.

Lab-scale electrocoagulation (EC) studies are valuable for understanding the fundamental removal mechanisms of contaminants; however, they often fall short in replicating the complexities of real-world applications during scale-up.⁵ Large-scale EC reactors differ significantly in terms of electrode performance, energy consumption, material usage, and flow dynamics, all of which influence the overall efficiency and sustainability of the process.^6,7 As a result, environmental and economic assessments conducted solely at the lab scale may not adequately capture the long-term operational costs, maintenance demands, and full life cycle impacts. Moreover, large-scale systems can address operational and maintenance issues that are rarely possible at the laboratory scale, leading to potential underestimation or misrepresentation of actual environmental burdens.⁷ Therefore, performing life cycle assessment (LCA) across different scales is essential to identify scale-dependent environmental hotspots, validate the scalability of sustainable claims, and understand trade-offs between process performance and environmental impact. A comprehensive techno-economic and life cycle assessment considering all stages of operation, from raw material input to final treated water, under different treatment capacities, is thus critical for assessing the environmental and economic friendliness of EC-based water treatment systems in real life applications.

In the literature, life cycle assessments have been performed at lab-scale EC reactors with evaluation of midpoint impact categories.^4,10 However, midpoint impacts are more related to elementary flows and do not directly reflect the actual damage to human health, ecosystems, or resource availability. In contrast, endpoint impacts aggregate these midpoint results to provide a more insightful understanding of the final damages, aiding in broader decision-making and policy evaluation.⁸ Evaluation of endpoint impacts from the ReCiPe method is crucial for understanding the overall benefits in avoiding the burden of disease to the population in terms of disability adjusted life years (DALYs) from the treatment of arsenic and fluoride-contaminated water. They assist in evaluating whether the water treatment process genuinely indicates health benefits (in terms of DALYs) or if the environmental costs exceed the advantages—this could compromise the technology's sustainability. In the literature, a life cycle assessment of arsenic and fluoride removal through an integrated electrocoagulation-based system, including the evaluation of endpoint damages, has not been reported.

Further, most of the economic analysis of the EC process has been performed by evaluating the operating cost only,^9,10 while only 3 studies have reported the CAPEX of the EC process.^11–13 Further, in OPEX, only raw materials and energy expenditure have been considered, avoiding the major contributions of manpower, maintenance and land cost. True economic feasibility of any process can be assessed considering all the capital (including machinery, building, infrastructure etc.) and operational expenses (labour, maintenance, land, loan repayment, raw materials, utilities) etc. Furthermore, profitability of the treatment process also needs to be evaluated for real field long term application of the process. To the best of the authors' knowledge, there is no study on the detailed life cycle and techno-economic assessment of higher scale EC based water integrated treatment process available.

The present study performs a detailed LCA of EC based water treatment of arsenic and fluoride contaminated groundwater at different scales (reactor volume 1.7 L, 20 L, 100 L, 650 L and 1200 L) of treatment operation. The midpoint and endpoint impacts associated with the complete treatment process from EC treatment, sludge filtration and treated water filtration are considered for the study. Further, scenario analysis has been performed considering the different electricity generation sources to determine potential of impacts reductions. Single parameter sensitivity analysis and Monte Carlo uncertainty analysis has been performed to evaluate the accuracy, reliability, robustness and applicability of the obtained environmental impacts. Detailed techno-economic analysis (TEA) has been performed to evaluate the economic performance of EC based treatment process. Information in terms of CAPEX, OPEX, cash flow, revenue and economic indicators (net present value (NPV), internal rate of return (IRR), payback period (PB), and profitability index (PI)) have been evaluated to assess the economic viability of the process. Further, different scenarios of hours of operations, sludge utilization and renewable energy utilization have also been considered to identify optimal setups. Economic sensitivity analysis has also been performed to evaluate the impacts of critical parameters on economic performance of the process.

2. Materials and methods

2.1 LCA framework, definition, and modelling assumptions

2.1.1 LCA goals and system boundary. The present study aims to evaluate and compare environmental emissions from arsenic and fluoride-contaminated groundwater treatment at different scales of the Al-electrode electrocoagulation process. First, the LCA model for the four EC reactor scenarios, i.e. R₁-lab scale (1.7 L; working volume 1.4 L), R₂-semi-pilot (20 L; working volume 18 L), R₃-pilot (100 L; working volume 80 L) and R₄-large scale (650 L; working volume 550 L), was developed using Sphera LCA for expert software (LCA FE version 10.9.0.31) for the treatment of arsenic (0.55 ppm) and fluoride (6 ppm) contaminated water (iron-free water). Further, LCA on the real field plant R₅-large scale (1200 L, working volume 1000 L) for the treatment of real field arsenic contaminated water, having co-existing iron, has been performed. These were modelled for ‘cradle to gate’ analysis. The operations of the treatment process included in the system boundary treatment process are the EC reactor, sedimentation tank and sludge filtration unit. It starts with mining, extraction and transportation of ores for aluminium sheet making to the final filtration of sludge. In the literature, for the sustainable management of the sludge, investigations have been made by solidification in bricks, utilisation as a catalyst support, etc. In the present study, the life cycle impacts of scale-up sizes were evaluated to treat a unit volume of synthetic groundwater. The amount of sludge does not vary much with different scales of the reactor. Thus, sludge management has not been included in the system boundary, since LCA has been done with a synthetic solution. Fig. S1 shows the system boundary diagram of the process.

2.1.2 Functional unit. This is the quality on which all assessed impacts were based. The functional unit for the present study is the treatment of 1000 L (1 m³) of groundwater having 0.55 ppm arsenic and 6 ppm fluoride. However, in a real sample, other ions are also present.

2.1.3 Life cycle inventory and modelling assumptions. The materials and energy inputs to the LCA model were adopted from different sources, such as the laboratory and field scale experimental database, the author's previous work (ref. 3 and the Indian patent IN 202311014986) and the Indian patent IN 202311014986) and other published literature, managed LCA content (MLC version 2024.2), etc. The German databases were adopted to develop the LCA model except for Truck. The global (GLO) database was used for Truck. Table S2 shows the inventory of materials and energy inputs to the LCA model.

2.1.4 Life cycle impact assessment methodologies. Environmental impacts from the life cycle of the treatment operation were evaluated based on the ReCiPe 2016 midpoint (H) method in 18 different matrices. The list of these categories has been provided in section S3 of the SI. Further, the ReCiPe endpoint (H) method was used to evaluate the endpoint damages to ecosystem, human health and resources.

The impact of temporal variations on climate change metrics, global warming potential (GWP) and global temperature potential (GTP) was measured for different time spans as per the recommendations of the Intergovernmental Panel for Climate Change (IPCC). The IPCC's 4th assessment report (2007) recommends evaluating GWP over 20, 100, and 500 years, and GTP over 50 and 100 years, to capture both short- and long-term climate impacts of greenhouse gases, helping policymakers make decisions for both near-term and long-term planning.^14–16

Single-parameter sensitivity and Monte Carlo uncertainty analysis were also performed to evaluate the impact of individual and simultaneous parametric variations, respectively, on the total environmental impact. The detailed methodologies are provided in section S4 of the SI.

2.1.5 Impact of distinct electricity sources. The impact of different electricity sources (both renewable and non-renewable) on the total environmental load was evaluated on the R₄-large scale (working volume 550 L) set-up. Four different electricity sources, namely, electricity from the grid mix, photovoltaic (solar), biogas, and wind, were assumed to fulfil the electricity demand of the process.

2.2 Techno-economic assessment

The techno-economic assessment (TEA) of the integrated electrocoagulation process has been performed by performing calculations in MS Excel for considering different scenarios. Different scenarios have been developed based on the number of operational shifts per day (each lasting 8 hours), sludge utilization (for catalyst support) and use of solar energy. The sludge generated in the present process isprimarily alumina-based. In the author's earlier work,¹⁷ this sludge was successfully utilized as a catalyst support material, demonstrating its technical feasibility for value-added reuse. Although large-scale market demand for such applications is still evolving, inclusion of sludge utilization in the TEA highlights the potential benefits of resource recovery and circular economy integration. Nevertheless, practical feasibility will depend on location-specific industrial linkages and regulatory acceptance. The details of scenarios are presented in Table 1. In the TEA, the costs of each process equipment have been evaluated based on the costs of materials and fabrication charges obtained from local vendors and literature. Table S3 of the SI presents the lists of economic parameters and assumptions used in the TEA, which was based on Indian costs (e.g. salary rates) and the published literature. The total capital expenditure (CAPEX), operating expenditure (OPEX), and revenue were calculated. The economics life of 20 years has been considered for the EC water treatment plant and the economic feasibility of the process was evaluated in terms of 4 indicators, namely, the net present value (NPV), internal rate of return (IRR), payback period (PB) and profitability index (PI). Detailed definition and mathematical expressions of these quantities are presented in section S5 of the SI. Further, a sensitivity analysis has been performed considering ±20% variations in the values of critical parameters and their effects on the economic indicators (NPV, IRR, PB and PI) were evaluated. Further, the detailed methodology including the mathematical equation and assumptions involved in the process are described in section S3 of the SI.

Table 1 Overview of techno-economic assessment (TEA) scenarios for electrocoagulation (EC) plant operations under varying shifts, sludge utilization, and solar energy utilization

S. no.	Scenario name	Description
1.	TEA-Sc.-100	EC plant operating in single (8 hours) shift
2.	TEA-Sc.-200	EC plant operating in two (16 hours) shifts
3.	TEA-Sc.-300	EC plant operating in three (24 hours) shifts
4.	TEA-Sc.-110	EC plant operating in single (8 hours) shift with sludge utilization
5.	TEA-Sc.-210	EC plant operating in two (16 hours) shifts with sludge utilization
6.	TEA-Sc.-310	EC plant operating in three (24 hours) shifts with sludge utilization
7.	TEA-Sc.-101	EC plant operating in single (8 hours) shift with obtaining electricity from solar energy
8.	TEA-Sc.-201	EC plant operating in two (16 hours) shifts with obtaining electricity from solar energy
9.	TEA-Sc.-301	EC plant operating in three (24 hours) shifts with obtaining electricity from solar energy
10.	TEA-Sc.-111	EC plant operating in a single (8 hours) shift with sludge utilization and obtaining electricity from solar energy
11.	TEA-Sc.-211	EC plant operating in two (16 hours) shifts with sludge utilization and obtaining electricity from solar energy
12.	TEA-Sc.-311	EC plant operating in three (24 hours) shifts with sludge utilization and obtaining electricity from solar energy

---

3. Results and discussion

3.1 Life cycle environmental impacts from different scales of operation

3.1.1 Relative contributions of distinct life cycle inputs to environmental metrics. The relative contribution of different materials and energy interactions are presented in Fig. 1 and 2. The contributions in GWP (kg CO₂ eq.) are majorly from the aluminium electrodes' dissolutions and electricity consumption for electrochemical reactions and stirring. The environmental impacts in the GWP category from aluminium are majorly attributed to the smelting process (∼71.3%) in aluminium production followed by alumina refining, ingot casting, anode production (for the Hall–Héroult process) and mining of bauxite ore.¹⁸ The major contributions from the smelting process of aluminium production are due to emissions of large quantities of methane, hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs) along with carbon dioxide, which have several times higher global warming potentials than the carbon dioxide emissions.^19,20 Further, fossil-based energy consumptions in the form of thermal and electrical energies also contribute to GWP due to emissions of carbon dioxide in different mining and extraction processes.²¹ Further, the electricity consumption for the electrochemical reaction and stirring (mixing) also contributed significantly to the EC-based treatment process. The aluminium transportation, sodium chloride addition and removal of arsenic from the water do not significantly affect the GWP from the process.

Fig. 1 The environmental impacts from the four scale scenarios of the EC process evaluated in different impact categories, (a) global warming potential (GWP, kg CO₂ eq.), (b) terrestrial acidification (TAP, kg SO₂ eq.), (c) particulate matter formation potential (PMFP, kg PM2.5 eq.), (d) ozone depletion potential (ODP, kg CFC-11 eq.), (e) freshwater consumption potential (FCP, m³), (f) land use (LUP, annual crop eq. y), (g) fossil depletion potential (FDP, kg oil eq.), (h) metal depletion potential (MDP, kg Cu eq.), and (i) ionizing radiation (IRP, kBq Co-60 eq. to air).

Fig. 2 The environmental impacts from the four scale scenarios of the EC process evaluated in different impact categories: (a) freshwater eutrophication potential (FEP, kg P eq.), (b) marine eutrophication (MEP, kg N eq.), (c) terrestrial eutrophication (TEP, kg 1,4-DCB eq.), (d) photochemical oxidant formation potential, human health (POFPH, kg NO_x eq.), (e) photochemical oxidant formation potential, ecosystem (POFPE, kg NO_x eq.), (f) freshwater ecotoxicity potential (FETP, kg 1,4-DCB eq.), (g) human toxicity potential, cancer (HTPc, kg 1,4-DCB eq.), (h) marine ecotoxicity potential (METP, kg 1,4-DCB eq.), and (i) human toxicity potential, non-cancer (HTP nc, kg 1,4-DCB eq.)

The particulate matter formation potential (PMFP) accounts for the emissions of particulate matter in the production processes. Primary aluminium production accounts for >60% PM emissions. The major emissions are from bauxite mining, alumina extraction, smelting and refining operation.²² Although most of the PM from mining is unspecified, the PM from the smelting process contains fluorinated compounds because of the presence of electrolyte NaAlF₄ in the Hall–Héroult electrolytic process.²³ The electricity consumption also has significant PM emissions due to the excavation of coal and its combustion at power plants.^24,25

Total electricity generation has a significant contribution (>80%) to the land use potential (LUP). It is due to direct and indirect occupation/transformation of agricultural land during the mining and electricity generation from coal. Lands utilized for mining have issues of land subsidence, soil degradation, land acidification, etc., making the land unsuitable for agriculture.^26,27

Marine and freshwater eutrophication potentials (MEP and FEP) are also mainly caused by electricity consumption (>80%). The higher contributions of electricity consumption for the MEP and FEP are caused by emissions of phosphate from fly and bottom ash to water booties, and emissions of nitrogen oxides in the combustion process of the electricity generation process.²⁸ Similarly, the fossil depletion potential (FDP) is caused by the depletion of fossil energy resources. Aluminium dissolutions and electricity consumption majorly contribute (37–60%) to the FDP because of the depletion of fossil resources consumed in thermal and electrical energy generation processes for different operations.^29,30

The removal of arsenic from groundwater avoids the marine ecotoxicity (METP), freshwater ecotoxicity (FETP), human toxicity cancer (HTc) and human toxicity non-cancer (HTnc) potentials to a great extent. Negative values of marine and freshwater ecotoxicity potentials from arsenic removal are due to avoiding the bioaccumulation of arsenic in marine and freshwater species such as fish and phytoplankton. It causes hyperglycaemia, depletion of enzymatic activities and various other acute or chronic toxicities to aquatic life.^31,32 The negative HTc signifies the reduction in risks of human cancers, while negative HTnc shows avoidance of non-carcinogenic diseases in humans, such as arsenicosis, cardiovascular and neurological system disturbance, etc.³³

Aluminium electrode dissolutions (55–76%) contributed significantly to the metal depletion potential (MDP) followed by electricity consumption for electrochemical (21%) and stirring (2.5–23%) operations. The MDP quantifies the depletion of finite metal resources and their future scarcities. Aluminium dissolution directly causes the decrease in metal resources, therefore it has the highest MDP contributions. The MDP of electricity is attributed to utilization of large quantities of metals for development/manufacturing of electricity generation facilities and machineries (such as turbines, copper wires in generators, etc.).³⁴

The terrestrial acidification potential (TAP) is significantly contributed by aluminium electrode dissolutions (60–80%) which is attributed to emissions of acidifying gases such as oxides of nitrogen and sulphur. Utilization of thermal and electrical energy resources in the alumina refining and smelting processes majorly contributes to the TAP. Further, the TAP from electricity consumption in electrochemical (17–18%) and stirring (2–20%) operations are due to emissions of NO_x and SO_x from combustion of fossil fuels at thermal power plants.³⁵ Similarly, terrestrial ecotoxicity potential (TETP) from aluminium electrode dissolution and electricity consumption is caused by the release of heavy metals and toxic compounds in the form of waste from mining and extraction processes.

The photochemical oxidant formation potential, human health (POFPH) and photochemical oxidant formation potential, terrestrial ecosystem (POFPE) represent the impact of photochemical smog like ground level ozone on human health and terrestrial ecosystems. The consumption of fossil fuels in aluminium production and electricity generation emits VOCs, NO_x, SO_x, CH₄ and CO, which forms these photochemical oxidants in the presence of sunlight.³⁶

The ozone depletion potential (ODP) quantifies the environmental impacts which contribute to the degradation of the ozone layer. Electricity consumption in electrochemical and stirring operations is the major contributor (60–80%) to the ODP. It is primarily due to emissions of GHGs like CO₂, CH₄ and N₂O to the atmosphere from the mining and transport to the electricity generation phase of fossil-based electricity production and also from emissions of halons 1211 and 1301 used as fire suppressants in mining operations.^37–39 The contribution to the ODP from aluminium electrodes (16–33%) is also attributed to electricity consumption in the mining, refining and smelting processes.

3.1.2 Environmental implications at different scale of operations. The comparative assessment of environmental implications from different scales of operations is presented in Fig. 1 and 2. Further, absolute values of environmental impacts for different reactor capacities are presented in Table S3 of the SI. It is evident from the figure that with an increase in the scale of operation, environmental implications in almost all the categories except ecotoxicities have shown significant reduction in impacts. For instance, total GWP decreases from 5.38 kg CO₂ eq. for 1.4 L reactor to 3.82 kg CO₂ eq. for 18 L, 2.98 kg CO₂ eq. for 80 L and 2.63 kg CO₂ eq. for 550 L. Similarly, other impacts, namely, FEP, ODP, LUP and MEP, decrease by 38–43% when scaling up from the laboratory 1.4 L to 18 L, by 58–63% for scaling to 80 L and by 61–66% for scaling to 550 L. The reduction in the impact with the increase in the scale of the reactor is mainly attributed to the significant reduction in the electricity requirement for electrochemical reactions and stirring. The scaling up of the EC process from the lab scale to the semi-pilot, pilot and large-scale operations significantly reduces the electrochemical resistance of the reactor and hence decreases the voltage and energy requirements. It is due to the increase in the electrode surface area that decreases the ohmic and faradaic resistances of the electrocoagulation process.⁴⁰ Further, the electricity requirements for providing perfect mixing in the EC reactor also contributed significantly to environmental emissions at a lower scale of operations (53%). However, at higher scales of operation, the mechanism of providing perfect mixing has been switched from magnetic stirring to water recirculation, which significantly reduces the electricity requirement with a gradual increase in the scale of operation. For instance, the GWP for the stirring operation decreases from 1.79 kg CO₂ eq. for 1.4 L to 0.1 kg CO₂ eq. for 550 L (94.28% reduction) and similarly for other impact categories such as the ODP, TETP, FDP, etc.

Moreover, a slight decrease in impacts has also been contributed to by a decrease in theoretical aluminium dissolutions. The GWP of aluminium electrode dissolution in 550 L is 1.57 kg CO₂ eq., which is ∼20% less than the GWP for the 1.4 L reactor system (1.96 kg CO₂ eq.). Furthermore, the negative value of total freshwater ecotoxicity, marine ecotoxicity, HTPc and HTPnc remains almost the same for all the reactor systems, since the amount of arsenic removed through the treatment process remains the same for each reactor system.

Similar reduction in the GWP, ODP, FETP, HTP and other environmental footprints with the scaling-up of the electrochemical oxidation process from a lab-scale (250 mL) reactor to a higher scale (1000 L) reactor has been reported in the literature.⁴¹ These reductions in environmental footprints are attributed to increased energy efficiency at a higher scale of operations. Further, mass transport and current distributions also have a paramount effect on the overall performance of the system. Similar reductions in environmental impacts with the scale-up have also been reported in the literature.^42,43

3.1.3 Endpoint impacts to ecosystem, human health and resources plots. Fig. 3 elucidated the endpoint impacts from the treatment of arsenic and fluoride-contaminated groundwater at different scales of operation. While the midpoint method is a problem-oriented impact assessment method which quantifies the emissions that affect a particular environmental category, the endpoint is a damage-oriented impact assessment method which aggregates the categories from the midpoint methods to provide the overall effect.⁴⁴ It quantifies the impact in three major categories, namely, the impact on human health burden, ecosystem and resources. It is evident from Fig. 3(a), that treatment and consumption of arsenic and fluoride contaminated groundwater can reduce the human health burden by making its impact value negative, calculated in units of DALYs (disability adjusted life years).⁴⁵ Further, treatment on a higher scale of operation improves the health benefits. The major contributor to the negative value of human health burden is HT nc, as evident from Fig. 3(b), which is due to the removal of arsenic from contaminated groundwater. The presence of arsenic and fluoride in the water causes human health issues, and thereby removing the arsenic reduces the overall human health burdens (damages).

Fig. 3 (a) ReCiPe endpoint impacts from the EC process, (b) contributions of different midpoint to three endpoint impacts, namely, ecosystems, human health burden and resource depletion impact categories [CC-TE: climate change-terrestrial ecosystems, FC-TE: freshwater consumption-terrestrial ecosystem, LO: land occupation, POF-E: photochemical ozone formation, ecosystems, TA: terrestrial acidification, CC-HH: climate change human health, PMF: particulate matter formation, MD: metal depletion, FD: fossil depletion], and (c) distribution of different process inputs to the most contributing midpoint categories.

The damage to the ecosystem, as evident from Fig. 3(a), reduces with an increase in the scale of operation from R₁ to R₄. This metric represents the loss of biodiversity due to environmental degradation, calculated in terms of species lost per year. The major environmental category contributing to the ecosystem damage is climate change-terrestrial ecosystem (CC-TE) (as shown in Fig. 3(b)). Further, the contributions of different materials and energy interactions to CC-TE are shown in Fig. 3(c). It is evident from Fig. 3(c) that the utilisation of aluminium electrodes contributes significantly to the damage of the ecosystem, along with the emissions from electricity consumption.

The depletion in overall resources is contributed by two environmental categories namely metal depletion and fossil depletion. The fossil depletion contributes to ∼98% of overall impact on resource depletion (Fig. 3(b)). Further, it is evident from the Fig. 3(c) that utilization of aluminium electrodes majorly contributes to the fossil depletion and hence overall damage to resources.

3.1.4 Impact of temporal variations on climate change metrics. The effect of temporal variations on the climate change metrics namely GWP (20, 50 and 100 years) and GTP (50 and 100 years) are presented in Fig. 4. The GWP is the measure of cumulative heat absorbed over a given time due to emissions of gases while the GTP is a measure of the change in temperature at the end of that period (relative to CO₂). It shows significant reduction in the impact with increase of timeframe. For instance, the GTP over 100 years is 0.9% to 0.55% less for 1.4 L (R₁) to 550 L (R₄) reactor systems than the GTP over 50 years. Similarly, the global warming potential (GWP) over a 100-year timeframe shows a reduction of 7.81% for R₁, 6.3% for R₂, 6.43% for R₃, and 6.4% for R₄ compared to the GWP over 20 years. Similarly, the GWP over a 500-year timeframe shows a reduction of 10.8% for R₁ and 8.45% for R₂, R₃, and R₄ compared to the GWP over 20 years. These reductions in emissions at higher time horizons are depicted to the less atmospheric lifetime of significant greenhouse gases (GHGs). For instance, methane has an average lifetime in the atmosphere of 8 to 12 years because methane decays into carbon dioxide (CO₂) with time.⁴⁶ Methane has ∼26-times higher GWP than CO₂, therefore, it helps in reduction of impacts over higher time horizons. Similarly, other GHGs also degrade over time. Thus, this suggests that the short-term impacts on the climate change are relatively greater than the long-term impacts. This is consistent with the other studies.

Fig. 4 Impact of temporal variation on the climate change metrics (a) GTP and (b) GWP.

3.1.5 Environmental implication from different electricity sources. Fig. 5 and S2 elucidate the environmental impacts from the electrocoagulation process (R₄-550 L reactor system) for different electricity sources, namely, electricity from grid mix, solar, biogas and wind. Further, absolute values of environmental impacts from different electrical energy scenarios are presented in Table S7 of the SI. The utilization of renewable energy sources such as solar, biogas and wind electricity reduced the GWP of the process by 36.7% 21.3% and 38.7%, respectively, compared to that of the grid electricity, which is majorly fossil based. The higher GWP from biogas based electricity is attributed to the emission of carbon dioxide (from combustion of methane) and methane (from uncombusted and leakage), whereas solar and wind based electricity are completely carbon free.⁴⁷ Further, utilization of solar based electricity significantly reduces the impacts in the FDP (36.3%), FCP (14%), POFPE (24%), POFPH (23%), ODP (63%), MEP (72.7%), FEP (79%), LUP (79.4%), MDP (16.1%), PMFP (17%), TAP (14.1%), and IRP (26.6%) compared to the impacts from utilizing grid electricity. Similarly, the use of wind electricity has significantly less impacts compared to grid mix electricity, for instance, FDP (40%), FCP (15.7%), POFPE (27.8%), POFPH (28%), ODP (64.3%), MEP (75%), FEP (81.1%), LUP (80.6%), MDP (14.6%), PMFP (17%), TAP (18.2%), and IRP (26.7%). These significant reductions in almost all categories from utilization of solar and wind energies are attributed to the avoidance of fossil-based electricity sources, which has a greater percentage in grid mix electricity.

Fig. 5 Environmental impacts from distinct electricity scenarios [Sc. 1: electricity from grid mix, Sc. 2: electricity from solar energy, Sc. 3: electricity from biogas, and Sc. 4: electricity from wind energy]: (a) GWP, (b) MEP, (c) ODP, (d) FCP, (e) LUP, (f) FETP, (g) FDP, (h) POFPE and (i) MDP.

In contrast, the electricity from biogas has shown a significant increase in the MEP (11 times), FEP (14 times), FCP (1.28 times), LUP (12.2 times), POFPE (1.57 times), POFPH (1.56 times), MDP (3.53 times), PMFP (1.46 times) and TAP (1.56 times) impact categories than grid electricity. The increase in the MEP and FEP are primarily attributed to the release of ammonia, nitrate and phosphates from the digested liquid used as fertilizer. Excess nutrients from this digested liquid leached out to water bodies. Similarly, the TAP is attributed to the release of NO_x and SO_x during the combustion of biogas.⁴⁸ The increased FCP is due to higher water requirements in the anaerobic digester. The significantly higher LUP is caused by increased occupied surface area by production of crops and biogas plants. The POFPE, POFPH and PMFP are attributed to the release of methane and other gases during combustion of biogas.⁴⁹ Further, the greater MDP is due to the utilization of metal-based equipment for handling of gas in biogas power plants.⁵⁰ These results are in good agreement with the previously reported studies.^51,52

3.1.6 Endpoint impacts from different electrical energy scenarios. Fig. 6 elucidates the endpoint impacts from the treatment of arsenic and fluoride contaminated groundwater at the 550 L electrocoagulation system with different electricity sources. It is evident from Fig. 6(a) that negative human health burdens in the DALYs for all the electricity sources show significant improvement in human health. However, utilisation of solar and wind energies is more beneficial for improving human health than the use of electricity from grid mix and biogas. It is due to having a higher magnitude of negative human health impacts in terms of DALYs. For instance, solar and wind electricity have −8.2 and −8.3 DALY values, whereas electricity from grid mix (−7.3 DALY) and biogas (−7.6 DALY) have less magnitude of negative human health impacts. Like different reactor scenarios, the major contributor to the negative value human health impacts is the HT nc category, as evident from Fig. 6(b), which is due to the removal of arsenic from contaminated groundwater (Fig. 6(c)) for all the electrical energy scenarios. It is also evident from Fig. 6(c) that for biogas, electricity production contributes to human health positively and hence makes more human health burdens compared to solar and wind energies.

Fig. 6 (a) ReCiPe endpoint impacts from the EC process for different electricity scenarios, (b) contributions of different midpoint to three endpoint impacts, namely, ecosystems, human health burden and resource depletion impact categories, and (c) distribution of different process input to the most contributing midpoint categories.

The aggregate impacts on ecosystems from different electricity sources show that electricity production from biogas has significantly contributed to ecosystem damage. Solar and wind electricity have significantly less (37.63% and 39.8%) ecosystem damage than grid mix electricity. The major contributor to ecosystem damage is climate change in the terrestrial ecosystem for all the electrical energy scenarios except electricity from biogas. For the electricity from biogas, land occupation or land use potential contributed significantly.

The resource depletion is also significantly less for solar and wind electricity scenarios. For instance, 16.4% and 17.2% reduction in resource depletion damage has been observed when using solar and wind electricity, respectively, in place of grid electricity. The fossil depletion contributes to 98% of the overall impact on resource depletion (Fig. 6(b)). Furthermore, Fig. 6(c) clearly indicates that the use of aluminium electrodes significantly contributes to fossil fuel depletion, thereby exacerbating overall resource depletion.

3.1.7 Life cycle sensitivity analysis. Sensitivity analysis was done to analyse the influence of ±20% variation in the values of certain parameters (such as amount of aluminium dissolutions, electricity for electrochemical reaction, electricity for stirring and Truck transportation distance) on the process environmental impacts. The variation of impacts in different impact categories are shown in Fig. 7 and S3. The aluminium dissolution is the most sensitive parameter across 10 impact categories, causing a total environmental impact variation of ±20%, with specific variations of the GWP (±10.9%), FDP (±12%), FCP (±17.7%), PMFP (±16.1%), TETP (±12.6%), TAP (±16%), POFPH (±14%), POFPE (±14%), MDP (±15.7%), and IRP (±14.1%). Following aluminium dissolution, the electricity consumption for electrochemical reactions has the largest influence in four environmental impact categories, namely, the ODP (±13%). FEP (±16.3%), MEP (±15.1%) and LUP (±16.3). Further, the variation in ±20% arsenic removal causes significant inverse variations in toxicities, for instance, HT c increases by 20.7% when arsenic removal decreases by 20% and similarly for HT nc (∓20%), FETP (∓20.1%) and METP (∓20.3%). Thus, the results of sensitivity analysis indicate that aluminium and the energy source for electrochemical operations must be efficient and precisely maintained. Further, arsenic removal efficiency should also be precisely observed to avoid the human, freshwater and marine toxicities.

Fig. 7 Impact of parametric sensitivity on the overall environmental impacts: (a) GWP, (b) METP, (c) MEP, (d) FCP, (e) PMFP, (f) TETP, (g) FEP, (h) TAP and (i) ODP.

3.1.8 Monte Carlo uncertainty analysis. The results of LCA may contain uncertainties due to unpredictable variation in input parametric values and errors in the calculations of material and energy data. Therefore, it is pivotal to consider and assess LCA uncertainty in order to enhance the credibility, acceptability, and reliability of LCA results.⁵³ Single-parameter sensitivity analysis shows the variation of impacts by varying an individual parameter, keeping other parameters constant, while Monte Carlo uncertainty analysis elucidates the variation in impacts by simultaneously varying all the parameters. The results of Monte Carlo uncertainty analysis in terms of means, medians, standard deviations and estimates at the 10th, 25th, 75th and 90th percentiles of the distribution are presented in Tables S8 and S9. The calculated mean and median values are identical to those of the basic scenario for both normal and uniform distributions. The standard deviation (<20%) in almost all the impact matrices for both normal and uniform distributions shows the acceptability of the results obtained from simultaneous variation of parameters for each scale of operations. Furthermore, the impact values at the 10th, 25th, 75th, and 90th percentiles of the distribution closely align with those of the basic scenario, demonstrating the robustness and reliability of the assessed environmental impacts.

A comparison of standard deviations and medians obtained through normal and uniform distributions is presented in Table S10. It is evident from the table that the standard deviations and medians are higher for the normal distribution than those for the uniform distribution. This indicates that while the choice of distribution influences the spread of results, the overall trends and conclusions of the study remain consistent, confirming the robustness of the LCA outcomes.

3.2 Life cycle impacts from the real field plant

The environmental impacts from the real field plant (R₅) are presented in Table S6. The results show significant reduction in environmental impacts with respect to (R₄) in all the impact categories. For instance, the GWP from treatment in R₅ (1.55 kg CO₂ eq.) is ∼40% lower than the GWP from treatment in R₄ (2.62 kg CO₂ eq.). These reduction in impacts are due to the presence of co-existing iron in the real groundwater significantly enhancing the removal efficiency of the arsenic and hence decreasing the electricity and aluminium requirements. Therefore, in real-field applications, large-scale EC operations exhibit significantly lower environmental impacts compared to lab-scale studies. Consequently, a techno-economic analysis (TEA) was subsequently carried out for the real-field scale system.

3.3 Techno-economic analysis

3.3.1 Capital and operating expenditures. The CAPEX and OPEX of different scenarios, along with the breakdowns of permanent equipment cost (PEC) and electricity consumption, are shown in Fig. 8. The CAPEX for all the base scenarios (namely TEA_Sc._100, 110, 200, 210, 300 and 310) are the same, i.e. INR 16.36 Lakhs. However, the CAPEX for the scenarios where solar energy has been utilised increases with the hours of operation. For instance, the CAPEX for the 8-hour, 16-hour and 24-hour operations with solar energy are INR 20.86 Lakhs, INR 23.86 Lakhs and INR 26.86 Lakhs. It is due to increased power requirements for 16- and 24-hour operations, which are higher than that for the 8 hour operation, therefore, a large CAPEX value is required for installing solar power systems, which adds significant cost in the CAPEX. The PEC accounted for 40–65% of the total CAPEX, equating to INR 10.66 Lakhs. For all the scenarios, the cost of DC power supply accounts for nearly 46.9% contributions to the total PEC, due to the requirement of high current for electrochemical treatment. Moreover, followed by the DC power supply unit, the settling tank accounted for 23.5% of the PEC which is due to fabrication of specifically designed stainless steel tank. The PEC of the EC_reactor (2.3%), EC_stand (3.5%), EC_wiring (2.3%), recirculation pumps (1.8%), NaCl dosing pump (0.9%), sand filters (3.5%) and overall piping system (3.5%) are relatively very less than the combined cost of the DC power supply unit and settling tank. Further, the costs of building, installation, utility set-up and other field expenses have much less contributions to the CAPEX relative to the PEC and solar power system cost.

Fig. 8 CAPEX and OPEX of the EC based water treatment process: (a) CAPEX for different TEA scenarios, (b) breakdown of permanent equipment costs (PEC), (c) OPEX for different TEA scenarios and (d) breakdown of electricity consumptions.

The OPEX of the water treatment plant shown in Fig. 8(c) depicts the variation in total operating cost with different scenarios. The OPEX of the process majorly depends on the raw materials, energy consumptions, manpower, land, maintenance and other costs. Since for different hours of operation scenarios, the treatment capacities are different, therefore, the operating cost for treatment of 1 mega liters (ML) of water is also shown in Fig. S4 of the SI. It is evident from Fig. 8(c) that the salary of manpower required for the operation of the EC treatment plant accounted for a major portion (46–56%) of the total OPEX. Further, the overall maintenance of the plant contributed for 20–23% costs of the total OPEX. The total electricity consumption accounted for 12–13% cost in the scenarios where grid electricity has been used, namely, TEA_Sc._100, 110, 200, 210, 300 and 310. The electricity consumption does not have any contribution to the OPEX in the scenarios where solar electricity has been utilized (TEA_Sc._101, 111, 201, 211, 301 and 311). The break-up of total electricity costs (Fig. 8(d)) shows that the cost of electricity consumption for electrochemical treatment has a significant contribution (57.4%). The costs of Al electrode, sodium chloride (NaCl) and maintenance of solar power systems have a 9–12%, 0.2% and 2.7–4.3% contribution to the total OPEX. Moreover, it is also evident from Fig. S4 that the OPEX per ML of water treatment slightly decreases (∼8.6%) for solar energy scenarios. Hence, the increased CAPEX for solar energy scenarios can be covered by reducing the OPEX electricity consumptions.

3.3.2 Cash flows. The results of cash flow analysis are shown in Fig. 9 and S5. The net cash flow was evaluated by considering the total sales, OPEX, depreciation, maintenance, loan repayment and tax of the process. The cash flow was higher for increased hours of operation, for instance, cash flows for the 20th year for 8 hours (TEA_Sc._100), 16 hours (TEA_Sc._100) and 24 hours (TEA_Sc._100) are INR 34.9, 68.5 and 100.8 Lakhs, respectively. This substantial rise in total cash flow with longer operating hours can be directly attributed to the enhanced water treatment capacity achieved through extended daily operation durations. Moreover, a slight increase in cash flows of 1.7–4.8% has been observed for scenarios where the cost of spent sludge has been included in sales. Furthermore, the cash flows for solar energy scenarios (TEA_Sc._101, 111, 201, 211, 301 and 311) increase by 6.8–7.8% with respect to their base scenarios (TEA_Sc._100, 200, and 300). This increase in cash flow for solar energy scenarios is due to the reduction in operating expenditure by avoiding grid electricity. It is also evident from the figures that the total sales and OPEX are the major contributors to the NCF in positive and negative directions, respectively.

Fig. 9 Cash flow and its breakdown into sales, operating cost, depreciation, tax and loan repayment: (a) TEA_Sc._100, (b) TEA_Sc._200 and (c) TEA_Sc._300.

3.3.3 Economic indicator performances. The economic performance (profitability) of different scenarios was assessed through four indicators, namely, the net present value (NPV), internal rate of return (IRR), payback period (PB) and profitability index (PI). The performance of the economic indicators is presented in Fig. 10(a) and (b). The results suggest that each of the scenarios is financially feasible. The minimum IRR of 19.5% and PI greater than 1 strongly suggest that the EC-based treatment has good economic potential. Further, the intermediate utilization of the generated sludge enhances the economic viability of the process. For example, the NPV and IRR for the sludge utilization scenarios (TEA_Sc._110, 210, and 310) are 7.5%, 3.1%, and 1.96% higher, respectively, compared to their corresponding base scenarios (TEA_Sc._100, 200, and 300). Similarly, the PB period of sludge utilization scenarios decreases by 2.2–6.3% with respect to their base scenarios while increasing the profitability index.

Fig. 10 Economic indicators of TEA: (a) NPV and IRR and (b) PB and PI for different TEA scenarios.

The use of solar electricity has a negative effect on the economic performance of the treatment process. The utilization of solar electricity slightly increases the NPV by 3.1–6.3% due to the reduction in the operating cost. However, the IRR decreases from 22.0%, 48.0% and 75.1% for base scenarios (TEA_Sc._100, 200, and 300) to 19.5%, 35.6% and 48.4% for their corresponding scenarios with utilization of solar electricity (TEA_Sc._101, 201, and 301), respectively. Similarly, the profitability index also decreases from 1.4, 1.93 and 2.46 for base scenarios (TEA_Sc._100, 200, and 300) to 1.34, 1.66 and 1.93 for their corresponding scenarios with utilization of solar electricity (TEA_Sc._101, 201, and 301), respectively. This decrease in the IRR and PI is attributed to the increased CAPEX from solar energy scenarios. Therefore, despite a moderate decline in economic indicators such as the IRR and PI due to higher capital investment, the adoption of solar electricity remains a favourable option considering its long-term sustainability and significantly lower environmental impact.

Further, although 24-hour operation scenarios have better economic performances, there are also some practical challenges involved with the process, for instance, difficulties in scheduling regular maintenance, night shifts and continuous sludge disposal, processing and management.

3.3.4 Economic sensitivity analysis. The effect of key parameters on the net present value (NPV) of the three base scenarios (TEC_Sc._100, 200 and 300) was examined. The fluctuation of ±20% in aluminium price, electricity unit price, manpower salary, interest rate and tax rate was considered as a key influencing parameter. Fig. 11 shows the effect of these parameters on the total NPV of the process. As evident from Fig. 11, the salary of manpower, followed by the tax rate, are the most sensitive parameters in the TEA of the treatment process. The ±20% variation in manpower and tax rate affects the NPV by ∓6.9–9.2% and ∓4.8–6.4%, i.e., an increase in manpower salary and tax rate significantly reduces the NPV and vice versa. Moreover, the variation in aluminium and electricity unit price has a very small effect on NPV (∓1.8–2.4%). Furthermore, variation in the interest rate has shown minimal variation in the NPV (∓0.57–2.22%). Hence, optimising manpower costs and tax policies emerges as a critical strategy to enhance the economic viability of the treatment process, while fluctuations in aluminium and electricity prices have comparatively minimal impact on overall profitability. The manpower cost estimations were based on Indian rates; nevertheless, the sensitivity analysis demonstrates that even with ±20% variation, the relative economic trends remain consistent, thereby facilitating comparison with international contexts. However, treatment costs may vary across different countries.

Fig. 11 Sensitivity of NPV of the EC treatment process economics with key process and economic parameters.

4. Conclusion

The life cycle assessment of the gradually scaled-up electrocoagulation-based arsenic and fluoride groundwater treatment system has been obtained, projecting the primary hotspot points of the emissions. The results of the present study depict significant reduction of environmental impacts for gradual scaled-up of the EC process from lab-scale (R₁) to large-scale (R₄). For instance, GHG emissions of the EC process reduced by 51% from R₁ to R₄. Further, aluminium dissolutions followed by electricity consumption in electrochemical reactions are the most critical contributing in the environmental emissions. Furthermore, treating real field water with co-existing iron showed lower (40%) environmental impacts compared to iron-free water. Sensitivity and Monte Carlo uncertainty analysis also affirmed the utmost impact of aluminium dissolution and electricity consumption. The endpoint impact analysis showed significant human health benefits for the treatment of arsenic and fluoride contaminated groundwater with some damages to the ecosystem and resources. The utilization of a carbon-free electricity source can further reduce the environmental emissions and improve the health benefits of the contaminated water treatment. The economic analysis, which calculated indicators, such as NPV, IRR, PB and PI, highlighted the financial viability of the arsenic and fluoride contaminated groundwater treatment through electrocoagulation approach. Increasing the hours of operation significantly improves the profitability of the treatment process. For example, the IRR values were 22.0%, 48.4% and 75.1% and NPV were INR 111.2 Lakhs, INR 273.5 Lakhs and INR 428.9 Lakhs for 8-, 16- and 24-hours operation base scenarios respectively. Although the CAPEX was higher for the solar energy scenario, t the OPEX is slightly lower compared to that for the grid energy scenario, making the scenario profitable. Moreover, base configuration was economically attractive for the short term, however for long term environmental sustainability solar energy scenario are recommended. Economic sensitivity analysis demonstrated that the manpower salary and tax rate are the most critical parameters in the TEA. Therefore, the study significantly assists in determining the environmental loadings of the arsenic and fluoride contaminated groundwater treatment system, together with suggesting different ways of making the technique more environment friendly at a large scale.

Author contributions

Hemant Goyal: conceptualization, methodology, software, validation, investigation, writing – original draft, visualization, and formal analysis. Prasenjit Mondal: conceptualization, supervision, and resources.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5ew00519a.

Acknowledgements

The authors are thankful to the Department of Science and Technology (DST), Government of India, for providing financial support (project number: DST-1553-CHD) and the Indian Institute of Technology Roorkee (India) for providing facilities and resources for carrying out this study.

References

1. https://www.who.int/news-room/fact-sheets/detail/arsenic, (accessed 15 February, 2025).
2. E. Shaji, K. V. Sarath, M. Santosh, P. K. Krishnaprasad, B. K. Arya and M. S. Babu, Geosci. Front., 2024, 15, 101734.
3. L. S. Thakur, H. Goyal and P. Mondal, J. Environ. Chem. Eng., 2019, 7, 102829.
4. H. Goyal and P. Mondal, Chemosphere, 2022, 304, 135243.
5. N. T. Thuy, N. X. Hoan, D. Van Thanh, P. M. Khoa, N. T. Tai, P. Q. H. Hoang and N. N. Huy, J. Electrochem. Sci. Technol., 2021, 12, 21–32.
6. S. Cotillas, J. Llanos, I. Moraleda, P. Cañizares and M. A. Rodrigo, Chem. Eng. J., 2020, 380, 122415.
7. M. C. Schulz, J. C. Baygents and J. Farrell, Int. J. Environ. Sci. Technol., 2009, 6, 521–526.
8. W. S. E. Ismaeel, J. Cleaner Prod., 2018, 182, 783–793.
9. J. Lu, Y. Wang, Y. Cao, Y. Liu, B. Lu, J. Xu, W. Wang and Z. Ni, J. Environ. Chem. Eng., 2025, 13(3), 116244.
10. S. M. Safwat, N. Y. Mohamed and M. M. El-Seddik, J. Environ. Manage., 2023, 328, 116967.
11. N. Anweshan, P. Mondal and M. K. Purkait, ACS ES&T Water, 2025, 5(5), 2591–2604.
12. R. F. Chen, C. H. Wei, H. T. Zhong, X. F. Ye, J. J. Ye, K. Liu, Q. B. Zhao and H. H. Ngo, Biosyst. Eng., 2024, 248, 47–57.
13. N. Karimi, S. A. Mirbagheri, R. Nouri and A. Bazargan, Results Eng., 2023, 17, 100770.
14. https://catalog.data.gov/dataset/ipcc-ar4-ar5-and-ar6-20-100-and-500-year-gwps#:~:text=TheInternationalPanelforClimate,com/USEPA/LCIAformatter, (accessed 15 February, 2025).
15. IPCC-2007, https://archive.ipcc.ch/publications_and_data/ar4/wg1/en/ch2s2-10-2.html, (accessed 21 February, 2025).
16. IPCC-2018, https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_Chapter08_FINAL.pdf, (accessed 21 February, 2025).
17. N. Lal, S. Gupta, H. Goyal and P. Mondal, Clean Technol. Environ. Policy, 2024, 26, 729–740.
18. M. Gautam, B. Pandey and M. Agrawal, Carbon footprint of aluminum production, Elsevier Inc., 2017.
19. B. K. Sovacool, S. Griffiths, J. Kim and M. Bazilian, Renewable Sustainable Energy Rev., 2021, 141, 110759.
20. R. Harvey, Estimates of U.S. Emissions of High- Global Warming Potential Gases and the Costs of Reductions, 2000.
21. E. Balomenos, D. Panias and I. Paspaliaris, Miner. Process. Extr. Metall. Rev., 2011, 32, 69–89.
22. S. H. Farjana, N. Huda and M. A. P. Mahmud, Sci. Total Environ., 2019, 663, 958–970.
23. N. P. Skaugset, D. G. Ellingsen, H. Notø, L. Jordbekken and Y. Thomassen, Environ. Sci.: Processes Impacts, 2015, 17, 578–585.
24. P. Lu, J. Wu and W. P. Pan, 2010 4th Int. Conf. Bioinforma. Biomed. Eng. iCBBE 2010, 2010, pp. 1–4.
25. N. Vig, K. Ravindra and S. Mor, Chemosphere, 2023, 341, 140103.
26. V. Fthenakis and H. C. Kim, Renewable Sustainable Energy Rev., 2009, 13, 1465–1474.
27. K. Ma, Y. Zhang, M. Ruan, J. Guo and T. Chai, Int. J. Environ. Res. Public Health, 2019, 16, 3929.
28. X. Cui, J. Hong and M. Gao, Energy, 2012, 45, 952–959.
29. B. Atilgan and A. Azapagic, J. Cleaner Prod., 2015, 106, 555–564.
30. Y. Zhang, M. Sun, J. Hong, X. Han, J. He, W. Shi and X. Li, J. Cleaner Prod., 2016, 133, 1242–1251.
31. B. Kumari, V. Kumar, A. K. Sinha, J. Ahsan, A. K. Ghosh, H. Wang and G. DeBoeck, Environ. Chem. Lett., 2017, 15, 43–64.
32. J. M. Neff, Environ. Toxicol. Chem., 1997, 16, 917–927.
33. U. Schuhmacherwolz, H. H. Dieter, D. Klein and K. Schneider, Crit. Rev. Toxicol., 2009, 39, 271–298.
34. J. Lieberei and S. H. Gheewala, Int. J. Life Cycle Assess., 2017, 22, 185–198.
35. A. Milovanoff, I. D. Posen and H. L. MacLean, J. Ind. Ecol., 2021, 25, 67–78.
36. P. Patel, I. Dimitriou, P. Mondal, O. Singh and S. Gupta, Energy Convers. Manage., 2025, 323, 119211.
37. B. Atilgan and A. Azapagic, J. Cleaner Prod., 2015, 106, 555–564.
38. E. A. Bouman, A. Ramirez and E. G. Hertwich, Int. J. Greenhouse Gas Control, 2015, 33, 1–9.
39. S. Schwietzke, O. A. Sherwood, L. M. P. Bruhwiler, J. B. Miller, G. Etiope, E. J. Dlugokencky, S. E. Michel, V. A. Arling, B. H. Vaughn, J. W. C. White and P. P. Tans, Nature, 2016, 538, 88–91.
40. Z. Gu, Z. Liao, M. Schulz, J. R. Davis, J. C. Baygents and J. Farrell, Ind. Eng. Chem. Res., 2009, 48, 3112–3117.
41. Y. Sun, S. Bai, X. Wang, N. Ren and S. You, Environ. Sci. Technol., 2023, 57, 1456–1466.
42. F. Piccinno, R. Hischier, S. Seeger and C. Som, J. Cleaner Prod., 2018, 174, 283–295.
43. S. Gavankar, S. Suh and A. A. Keller, J. Ind. Ecol., 2015, 19, 51–60.
44. https://www.sipl-sustainability.com/life-cycle-impact-assessment-impact-categories, (accessed 25 February, 2025).
45. https://www.who.int/data/gho/indicator-metadata-registry/imr-details/4657#:~:text=Disability-adjustedlifeyears(DALYs)attributabletothe environment(%25), (accessed 28 February, 2025).
46. M. L. Fagan, J. Bus. Strategy, 1991, 12, 21–25.
47. M. Granovskii, I. Dincer and M. A. Rosen, Int. J. Hydrogen Energy, 2007, 32, 927–931.
48. O. Hijazi, S. Munro, B. Zerhusen and M. Effenberger, Renewable Sustainable Energy Rev., 2016, 54, 1291–1300.
49. L. Lijó, S. González-García, J. Bacenetti, M. Fiala, G. Feijoo, J. M. Lema and M. T. Moreira, Renewable Energy, 2014, 68, 625–635.
50. C. Xu, W. Shi, J. Hong, F. Zhang and W. Chen, Renewable Sustainable Energy Rev., 2015, 49, 169–177.
51. M. Liu, A. Ogunmoroti, W. Liu, M. Li, M. Bi, W. Liu and Z. Cui, Sci. Total Environ., 2022, 807, 150751.
52. A. Whiting and A. Azapagic, Energy, 2014, 70, 181–193.
53. Z. Sheikholeslami, M. Ehteshami, S. Nazif and A. Semiarian, Process Saf. Environ. Prot., 2022, 168, 928–941.

---