Carla
Caldeira‡
,
Elisabetta
Abbate
,
Christian
Moretti
,
Lucia
Mancini
and
Serenella
Sala
*
European Commission, Joint Research Centre, Via Enrico Fermi 2749, Ispra, Italy. E-mail: serenella.sala@ec.europa.eu
First published on 31st May 2024
In the context of the EU Chemicals Strategy for Sustainability, a key action regards the development of a framework to identify criteria for safe and sustainable by design chemicals and materials. The integration of safety and sustainability considerations is challenging, and this systematic review investigates how aspects pertaining to sustainability have been implemented in 155 frameworks proposed by scholars, industry, governments and non-governmental organizations. In particular, this review scrutinizes methods, models and indicators for environmental, social and economic aspects in frameworks combining multiple sustainability dimensions. Furthermore, the application of such frameworks to an early stage of chemicals and materials development was also analysed. The review unveiled that the majority of the frameworks are purely conceptual/theoretical, while some attempts are made by others towards providing methods and indicators for the assessment as well as operational procedure of decision support. Life cycle considerations are often remarked as necessary for evaluating the environmental sustainability of chemicals, climate change being the environmental impact mentioned by the majority of frameworks. Social sustainability aspects with quantitative indicators have been proposed only in a few studies so far. Another aspect often disregarded is data uncertainty. Although the reviewed frameworks showed several similarities in structure and aspects covered, indicators often differ significantly. Hence, using one framework instead of another might lead to a different outcome.
A key policy goal defined in the European Green Deal is a zero pollution/toxic-free environment, together with climate neutrality, biodiversity protection, and circular economy.5 To support such ambition, the EU Chemicals Strategy for Sustainability (CSS) – Towards a Toxic-Free Environment puts forward actions to reduce impacts on human health and the environment associated with chemicals, materials, products, and services.6 In particular, the EU CSS calls for the definition of criteria for Safe and Sustainable by Design (SSbD) chemicals and materials by integrating safety, circularity and functionality, minimizing their life cycle environmental footprint.
The selection of safer alternatives has been the subject of several studies proposing frameworks for the assessment.7–9 Following these frameworks, viable or new alternatives are screened before commercialization to avoid regrettable substitutions. These frameworks include a hazard and risk assessment10–13 focusing mostly on the physicochemical properties (e.g. flammability), human toxicity (e.g. carcinogenicity) and ecotoxicity (e.g. bioaccumulation) of chemicals and materials. Within the European Union, environmental, health, and safety (EHS) legislation criteria are set by the REACH regulation.14
A seminal approach to considering sustainability aspects in chemical development was proposed in the field of green chemistry. The Green Chemistry concept was introduced in the environmental protection strategy of the U.S. (United States) Environmental Protection Agency (EPA) in the early 1990s.15 This concept then became well known with the publication of the 12 Green Chemistry principles by Anastas and Warner16 which consider efficient utilization of raw materials and elimination of waste and toxic and/or hazardous substances.17
In the past decade, the integration of sustainability aspects in the selection of chemicals and materials has been gaining prominence with the ambition of moving towards safer and more sustainable chemicals and materials.9,18–22 Incorporating chemicals’ sustainability aspects besides safety allows accounting for trade-offs between exposure of humans and ecosystems and environmental impacts (e.g. climate change) associated with chemical production and supply chains.
To gain insights into which safety and sustainability aspects would be relevant to be included in a framework for the development of SSbD criteria for chemicals and materials, the European Commission Joint Research Centre (EC-JRC) carried out an initial review on how sustainability aspects have been implemented in decision frameworks for safety, identifying which dimensions, aspects, methods and indicators have been proposed, as well as the decision approaches applied in the overall sustainability assessment framework.23 This review informed the development of the SSbD framework by the EC-JRC24 that underpins the EC Recommendation establishing a European assessment framework for safe and sustainable by design chemicals and materials.25 The framework considers the Green Chemistry principles key to design SSbD chemicals but their performance should be assessed by means of comprehensive sustainability assessment that considers the entire life cycle. A testing period by stakeholders is taking place and the revision of the framework built based on the feedback obtained during this period is foreseen. To inform the further development of the EC framework, a more systematic analysis of the frameworks is needed, especially unveiling the key scientific underpinning of the proposed framework, their level of operationalization and the focus to design support versus a proper and comprehensive assessment of the alternatives. Hence, this study aims to investigate indicators with respective methods covering sustainability aspects in frameworks integrating multiple sustainability dimensions and discuss the level of integration reached so far, highlighting frameworks used for the design of chemicals and materials, including in the early stage of development.
An overview of sources used to perform this review is presented in Fig. 1. The review builds and expands on the review carried out by the EC-JRC,23 the EC-RTD mapping study26 and the results of a targeted stakeholders’ survey.27 The latter provided information mainly on grey literature and existing legislation that considers sustainability aspects.
Fig. 1 Sources for the frameworks considered in this review. *as in Caldeira et al.23 updated to May 2023. **new query considered for this review. |
Moreover, the scientific literature obtained from the Scopus database with query 1 characterized by terms linked to the concept of safe and sustainable chemicals§ was updated in May 2023 including 868 articles. The search string used in Scopus was characterised also by the terms “solvent”, “selection” and “guide” since solvent selection guides reporting alternative assessment frameworks for solvent selection have been used for more than 20 years in the pharmaceutical sector,28 making this term well established. Moreover, since multi criteria decision analysis (MCDA) has been highlighted as a key instrument for sustainability assessment in general, as discussed in major works and reviews (e.g. ref. 29 and 30) an additional search in the Scopus database with query 2¶ was done, returning over 1400 results.
Once duplicates (studies captured in both reviews) were eliminated, the abstracts were revised. Those dealing with topics not related to chemicals and materials e.g. in supply chain management (e.g., supplier selection, transportation, location) and waste management (recycling, materials recovery, remediation) were excluded. The main text was considered in cases when reading the abstract and the title was not sufficient for such a screening. Additional frameworks not found directly by the Scopus search but cited by excluded case studies or reviews were also included in our analysis. In the end, 155 documents were considered in this review.
i. Coverage of sustainability dimensions (i.e. safety, environmental, social, and economic) and aspects (e.g. climate change) as well as which indicators and respective methods are suggested.
ii. Adoption of a life cycle approach and if so, what is the methodology and which are the environmental impacts considered. As mentioned in the Introduction, the EU CSS31 recalls the importance of a life cycle perspective in minimizing chemicals’ potential impacts to detect shifts in burdens between impact categories, life cycle stages or geographic locations.
iii. Decision support procedure implemented, including eventual scoring systems and the level of aggregation of the evaluation outcome as well as how data gaps and uncertainty were taken into account in the assessment.
Fig. 2 depicts the dimensions covered i.e. safety, and environmental, economic, and social sustainability. Despite this division, the authors recognize that safety is integrated in sustainability: it is important to note that safety is a wide concept embedded in several Sustainable Developments Goals, and chemical safety is stated in several targets relating to human health, environmental quality, and access to services and resources. However, since the SSbD concept distinguishes the two terms (safety and sustainability), the same was done in this work.
Most of the frameworks consider environmental and economic aspects47 or only environmental.37 A lower number of frameworks consider social aspects that are either combined with environmental aspects7 or with environmental and economic.20 For the latter in which environmental, economic and social aspects are considered there are frameworks suggested for example in ref. 34, 36, 39 and 49–54.
The work by Rossi et al.53 is a seminal framework to guide alternative assessment of chemicals, materials, and products. The evaluation performed via this framework look at four major areas: (1) impacts on human health and the environment, (2) social justice impacts, (3) technical performance and (4) economic feasibility. The framework proposed by the German Environmental Agency (UBA) investigates potential impacts of chemicals on human health and the environment and on social responsibilities in supply chains,34 while economic aspects are addressed to a minor extent only. The framework proposed by CEFIC (European Chemical Industry Council) is comprehensive in terms of covered dimensions but it is still at a conceptual level. The framework from the World Business Council on Sustainable Development (WBCSD) regards the Chemical Industry Methodology for Portfolio Sustainability Assessments (PSA).55 This framework is built on two established guidance documents on assessing environmental and social impacts of chemical products based on a life cycle approach.56,57 This framework is a major reference for frameworks further developed by companies implementing in-house PSA methodologies.51,54,58 The framework proposed by the Interstate Chemicals Clearinghouse (IC2) includes considerations of the full life cycle of the product. Both environmental and social impacts are considered via a set of modules. A priority is given on the modules regarding hazard, cost, availability, performance evaluation, and exposure assessment – whilst others (Materials Management, Social Impact, and Life Cycle) should be considered if relevant to the particular chemical, product, or process under assessment. The Cradle to Cradle Certified® Product Standard presents a list of requirements that products should comply with, ranging from human health to product circularity, climate protection, and social fairness. It also includes water and soil stewardship, general requirements and recommendation for packaging.
A different distribution of the dimensions included has been observed for the frameworks that take the early stage of development into account. Fig. 2 clearly shows the low percentage of frameworks applied in the early stage of development analysing social aspects. Conversely, the economic aspect is covered by 53% of these frameworks, a percentage that drops to 32% for frameworks that do not focus on the design phase. This suggests the central role of economic aspects when dealing with new chemicals or materials in order to assess further efforts in the development of the chemical/material under consideration. A similar trend can be observed for frameworks dealing with the environmental aspect, which is included in the majority of frameworks (around 80%), both those considering the early stage of development and those not. Frameworks focusing on single sustainability dimensions were rarely identified, suggesting that frameworks covering more than one aspect are preferred as they provide a broader analysis.
i. energy, including energy consumption/efficiency of a process or over the life cycle;
ii. circularity, considering features linked to reducing, reusing, repairing, refurbishing, remanufacturing, and recycling options;
iii. biodegradability, referring to the capacity for biological degradation of organic materials by living organisms down to the base substances such as water, carbon dioxide, methane, basic elements and biomass; and
iv. aspects related to the type and quantity of resources used and efficiency of the production process.
The indicators suggested in the frameworks associated with each aspect are summarized in Table 2, highlighting indicators used in frameworks in the early stage of development. Definitions and assessment methods are reported in the ESI.† Most of the frameworks include indicators related to the type and quantity of resources used. Indicators on biodegradability are seldom used, however with a slightly higher percentage in the case of frameworks considering the design phase. In contrast, circularity is highly analysed.
Aspect | Number of frameworks adopted | Early stage application | Aspect | Number of frameworks adopted | Early stage application |
---|---|---|---|---|---|
Indicator | Indicator | ||||
Resource, and processing- and product-related | 375 | 149 | Marine biodegradability [–] | 2 | |
Energy | 87 | 38 | Octanol–water distribution coefficient [Kow] | 2 | 2 |
Cumulative energy demand [MJ] | 17 | 12 | Soil biodegradability [–] | 1 | |
Energy conservation [–] | 2 | Resources: types, quantity, and efficiency considerations | 157 | 68 | |
Energy consumption [kW h or MJ] | 39 | 14 | Amount of (solid/water) waste [kg or %] | 17 | 4 |
Energy efficiency [%] | 17 | 4 | Atom economy [%] | 5 | 6 |
Energy intensity [kW h kg−1 or MJ kg−1] | 2 | 1 | Biomass consumption [kg] | 1 | |
Non-renewable resources with energy content [MJ] | 1 | Carbon economy [–] | 1 | 1 | |
Number of (process) steps [–] | 1 | 1 | E-factor [%] | 6 | 4 |
Primary energy demand [MJ] | 1 | 1 | Fossil fuel consumption [kg] | 1 | 1 |
Process carbon footprint | 1 | 1 | Global material economy [%] | 1 | 1 |
Process efficiency [%] | 1 | 1 | Hazard Waste [kg] | 2 | 1 |
Reaction efficiency [%] | 2 | 2 | Imported resources [E] | 1 | |
Recycling energy [E] | 1 | Mass yield [%] | 2 | 2 | |
Renewable resources with energy content [MJ] | 1 | Material conservation [E] | 2 | ||
Stoichiometric factor | 1 | 1 | Material intensity index [%] | 4 | 1 |
Circularity | 112 | 36 | Net mass of materials consumed [kg] | 6 | 4 |
Aqueous waste valorisation [E] | 1 | 1 | Non-renewable resource amount [kg] | 4 | 2 |
Boiling temperature [°C] | 10 | 5 | Non-hazardous waste [kg] | 2 | 1 |
Disassembly/reparability design [E] | 10 | Number of solvents [E] | 1 | 1 | |
Durability [years] | 13 | 4 | Raw material consumption [kg per ton] | 2 | 2 |
Durability of the building [years] | 1 | Raw material origin [E] | 2 | 2 | |
Energy requirement for recycling [MJ kg−1] | 2 | 1 | Reaction efficiency [%] | 10 | 5 |
Heat of vaporisation [MJ kg−1] | 2 | 1 | Recycled input materials [%] | 1 | |
Number of carbon atoms [E] | 1 | Relative process greenness [–] | 1 | 1 | |
Percentage of reclaimed products and their packaging materials [%] | 1 | Renewability of resources [%] | 5 | 5 | |
Purity of recovered solvent [%] | 2 | 1 | Renewable or fossil? [–] | 18 | 6 |
Recyclable?[–] | 12 | 3 | Renewable resource amount [kg] | 4 | 2 |
Recyclability/circularity [–] | 21 | 8 | Resource consumption [kg] | 4 | 2 |
Recycled content [%] | 12 | 1 | Resource efficiency | 1 | 1 |
Recycling efficiency/recovery rate [%] | 9 | 4 | Resource valorisation [–] | 2 | 2 |
Reuse rate/reusability [–] | 4 | 2 | Solid waste [kg per ton] | 2 | 1 |
Solvent selectivity [–] | 2 | Solid waste generation [kg] | 5 | 2 | |
Used organic solvent valorisation [–] | 1 | 1 | Solvent selectivity [–] | 1 | 1 |
Waste reduction [–] | 1 | 13 | Use of critical raw materials?[–] | 5 | 1 |
Waste utilization [–] | 5 | 3 | Waste characterisation potential [kg] | 1 | |
Yield of extraction (%) | 2 | Waste cleaning [–] | 2 | ||
Biodegradability | 19 | 7 | Waste reduction algorithm [–] | 1 | 1 |
Biodegradability [–] | 4 | 1 | Water consumption [m3] | 32 | 7 |
Biodegradability requirement [–] | 6 | Water efficiency [m3] | 1 | ||
Biodegradability [–] | 4 | 4 |
Various indicators measure the performance during distillation processes mostly found in frameworks applied to solvents as distillation is the major technique used by the chemical industries for recycling solvents.63 These indicators were proposed in terms of amounts (e.g. the energy requirement for recycling), efficiencies (e.g. yield of extraction) or physical properties (e.g. boiling temperature). However, environmental trade-offs of recycling chemicals are overlooked using these indicators. For example, the energy needed for distillation to recover a solvent might be higher than producing it.63
To minimize undesirable trade-offs of circularity, two innovative approaches were proposed in ref. 2 and 64. Chavarrio et al.64 proposed a quantitative multi-criteria decision method based on both the solvent and extraction processes under consideration. This method relies on criteria such as the cost of the solvent, yield of extraction, purity of recovered solvent, heat of vaporization, boiling temperature, solvent selectivity, etc. Wang and Hellweg2 proposed a two-step circularity assessment to evaluate approaches to reduce the major causes of chemical losses and qualitatively catalogue chemicals into six major categories leading to different management practices for recovering the embedded raw materials. As pointed out by Wang and Hellweg,2 most of the indicators used for assessing circularity are mass-based and can be misleading in guiding environmental sustainability. The authors give the example of lithium-ion batteries for which higher energy consumption and air pollution arises from current recycling technologies than from primary production. It is therefore essential to couple mass-based circularity indicators with methods that assess the environmental impacts of the ‘circular’ system.
As an environmental aspect, it was mostly qualitatively addressed. So, most frameworks mention biodegradability as an aspect causing environmental issues but not providing information on a specific indicator or method to be used. Indicators regarding biodegradability were often discussed with respect to the biodegradability of plastics based on specific standards e.g. ASTM D-640068 or the standard EN13432.69 In particular, no life cycle-based indicator regarding plastic littering was found. In fact, modelling littering requires a wide range of data regarding fate, exposure and effect modelling, which are mostly unavailable e.g. data regarding degradation rates of additives, effects from ingestion of plastic particles, etc.70 However, the LCA community is developing research in this field on the development of harmonized pathways to account for impacts of plastic litter, specifically to the marine environment.71
Water is the resource that most of the frameworks pointed out as a key aspect to consider, water consumption (m3) being the most recommended indicator at the pressure level (32 frameworks).
The amount of waste generated and the net mass of materials consumed were also recommended by 24 and 6 frameworks, respectively. The argument in favour of easy-to-calculate mass-based metrics measuring waste generation is often a proxy for the trends of most environmental impacts.4,17 Indicators typically used in Green Chemistry such as atom economy and E-factor and similar mass-based metrics that can be expressed in terms of E-factors (e.g. mass intensity = E-factor + 1)17 were also often proposed.
The idea behind mass-based metrics used in green chemistry is that the amount of waste generated is often a good proxy for all other environmental impacts.4 However, this assumption would lead to misleading outcomes in other cases, such as environmental comparisons between fossil and bio-based alternatives.17,22
Aspect | Number of frameworks adopted | Early stage application | Aspect | Number of frameworks adopted | Early stage application |
---|---|---|---|---|---|
Indicator | Indicator | ||||
Environmental | 459 | 181 | Photochemical Ozone Formation | 29 | 15 |
Pressure | Photochemical oxidant formation and ecosystems [kg NOx-eq to air] | 1 | 1 | ||
Air emission | 25 | 12 | Photochemical oxidant formation and human health [kg NOx-eq to air] | 1 | 1 |
Air emissions | 3 | 3 | Photochemical oxidation potential [PCO] | 2 | 2 |
Atmospheric CO2 emissions [g] | 3 | 1 | POCP [kg of ethene-eq or kg NMVOC or kg O3] | 25 | 11 |
Atmospheric NOX emissions [g] | 5 | 2 | Acidification | 35 | 12 |
Atmospheric SO2 emissions [g] | 3 | 1 | Acidification potential [molc H + eq] | 20 | 3 |
Atmospheric VOC emissions [g] | 2 | Aquatic acidification [kg H + ions] | 1 | 1 | |
Critical air mass [%] | 2 | 2 | Atmospheric acidification [kg SO2eq] | 2 | 2 |
Dimethylformamide [g] | 3 | 2 | Terrestrial acidification [kg SO2eq] | 12 | 6 |
Fine particulate matter (mg PM 2.5) | 4 | 1 | Eutrophication | 36 | 14 |
Water emission | 18 | 13 | Eutrophication potential [kg of phosphate ] | 19 | 5 |
Biological oxygen demand 5 [g] | 3 | 2 | Freshwater eutrophication potential [kg P eq] | 10 | 5 |
Chemical oxygen demand [g] | 3 | 2 | Marine eutrophication potential [mol N eq] | 6 | 4 |
Contaminant emission [g] | 1 | Terrestrial eutrophication potential [kg of N equivalents] | 1 | ||
Critical water mass [%] | 2 | 2 | Resources | 62 | 22 |
Total organic carbon [kg] | 4 | 4 | Damage to Resources [–] | 4 | 3 |
Water Discharge quality | 2 | Eco-indicator 99 depletion of resources [–] | 1 | 1 | |
Water emissions [kg m−3] | 3 | 3 | Fossil resource scarcity [kg oil eq.] | 2 | 1 |
Soil emission | 1 | 1 | Land occupation [m2a crop eq.] | 8 | 3 |
Soil emissions | 1 | 1 | Land requirements | 1 | |
Midpoint | Land Usage [M ha] | 1 | |||
Toxicity | 61 | 25 | Land use [kg C deficit] | 2 | 1 |
Bioconcentration factor | 1 | 1 | Land use [–] | 5 | |
Ecotoxicity [CTU] | 3 | 1 | Limiting depletion of fossil/mineral resources [–] | 4 | 2 |
Ecotoxicity [–] | 5 | Mineral resource scarcity [kg Cu eq.] | 1 | ||
Ecotoxicity damage [PDF m3d] | 1 | Resource depletion potential [kg Sb eq.] | 7 | 1 | |
Freshwater [eco]toxicity [CTUe] | 4 | 3 | Resource use/depletion – fossil [MJ] | 14 | 7 |
Freshwater ecotoxicity [g C6H4CI2-eq] | 5 | 2 | Water consumption potential [m3eq] | 1 | 1 |
Human toxicity – cancer potential [CTUh] | 4 | 2 | Water depletion potential [m3eq] | 10 | 2 |
Human toxicity – no cancer potential [CTUh] | 3 | 1 | water footprints [m3eq] | 1 | |
Human toxicity [–] | 7 | 1 | Endpoint | ||
Human toxicity potential [kg 1,4 DCB eq or CTUh] | 9 | 5 | Air pollution – air pollution and acidification [DALY] | 1 | 1 |
Indoor air quality [–] | 1 | Biodiversity protection (conceptual) | 10 | 1 | |
LCT human toxicity [mg intake] | 3 | 1 | Chemicals safety | 2 | 1 |
Marine ecotoxicity [g C6H4CI2-eq] | 3 | 2 | Climate change damage [DALY] | 2 | |
Respiratory inorganics | 1 | 1 | Air and water pollution [DALY] | 1 | |
Smog potential [kg NOx eq.] | 1 | Eco-indicator 99 total [–] | 8 | 2 | |
Terrestrial ecotoxicity [g C6H4CI2-eq] | 10 | 5 | Eco-indicator 99 ecosystem quality [–] | 2 | 1 |
Climate Change | 95 | 33 | Eco-Indicator 99 human health [DALY] | 2 | 1 |
C factor | 1 | Ecosystem quality (recipe) | 3 | 2 | |
Climate change [–] | 4 | 3 | Ecosystems (species year) | 5 | 1 |
CO2 balance | 1 | 1 | Ecosystem damage (conceptual) | 2 | |
GWP 100 [kg CO2eq] | 89 | 29 | Embodied biodiversity footprints | 1 | |
Stratospheric ozone depletion | 26 | 12 | Human health (DALY) | 8 | 2 |
Ozone depletion [–] | 1 | 1 | Human health (recipe) | 3 | 2 |
Ozone depletion potential [kg CFC-11] | 25 | 11 | Influence/impact on public health | 3 | |
Particulate Matter | 10 | 5 | Inherent safety indicator | 1 | |
Particulate matter [PM2.5eq] | 10 | 5 | Ionizing radiation-human health [DALY] | 1 | |
Ionizing radiation | 2 | 2 | Oral toxicity [−logLD50] | 1 | 1 |
Ionizing radiation [kBq U235 eq.] | 2 | 2 | Resource depletion [$] | 3 |
Regarding indicators at the pressure level, NOX, SO2, CO2 and fine particulate matter released into the air have been proposed as indicators in 3–4 frameworks. The indicator named the “critical air mass” index is an old environmental indicator proposed already in the nineties in chemicals’ selection tools and early studies on green chemistry.75,76 This indicator represents the mass of a specific type of air emissions (e.g. SO2) emitted by a process over a standard value typically representing the maximum acceptable amount of that pollutant e.g. based on legislation requirements.
Critical water mass is analogous to the critical air mass indicator but for the water compartment.75,76 Another indicator related to water emissions reported in the reviewed frameworks was the total organic carbon proposed to measure water pollution, especially by guides from the pharmaceutical industry.73,74,77 This indicator refers to the total soluble and insoluble organic matter entering water bodies. Biological oxygen demand and chemical oxygen have been proposed by three frameworks each. These indicators are common indicators to measure water quality.78 Finally, one indicator regarding soil emission was found at the pressure level mentioned by one framework.79 Still, the indicator was just mentioned without providing further information. Besides “soil biodegradability” in the resource dimension, no other indicators to evaluate impacts on soil were found. Indicators linked to the soil impacts are “Terrestrial eutrophication potential”, “Terrestrial ecotoxicity”, and “Terrestrial acidification”. This emphasizes the lack of interest in monitoring and assessing the impact on soils, focusing more attention on water and air emissions. In light also with the recent Proposal for a Directive on Soil Monitoring and Resilience,80 there is a need to develop and include indicators representative of the soil compartments both at the pressure and impact levels in the future.
A particular indicator with a single occurrence81 is a pressure-based indicator used as a proxy for toxicity in the EPA's GREENSCOPE (Gauging Reaction Effectiveness for the Environmental Sustainability of Chemistries with a multi-Objective Process Evaluator) tool. The definition of this indicator is detailed in Ruiz-Mercado et al.82 Essentially, this indicator represents the ratio between the total mass of toxins released over the total mass of products.
Regarding the environmental impacts, there were numerous (slightly) different versions of mid-point level indicators covering the same aspect in the reviewed framework. Therefore, indicators were grouped with the closest indicator in the counting when reasonably connected to another one in the list. For example, if the acidification potential was reported with a missing unit or a slightly different name, it was counted as the same indicator.
As presented in Table 3, environmental impacts at the midpoint level in the reviewed frameworks relate to climate change, toxicity, pollution such as acidification, eutrophication, ozone depletion, photochemical ozone formation, particulate matter and ionizing radiation, and resources and e.g. land use or water depletion.
The most suggested indicator in the frameworks is by far the global warming potential.89 This indicator represents the sum of Greenhouse Gas (GHG) emissions multiplied by the specific characterization factor. The calculation of this indicator depends on the time-scale, which was always 100 years in the reviewed frameworks reporting this indicator. The high occurrence of this indicator is due to both a broad acknowledgment of the priority of dealing with climate change and the scientific consensus on the model underpinning this indicator.83 This indicator is also adopted in various EU policies, especially for energy and alternative fuels.84,85
Stratospheric ozone depletion potential is suggested in 25 frameworks. The calculation of this indicator is based on a steady-state ozone depletion potential model.86 The indicator obtained from this characterization model represents the relative measure of the expected impact on ozone per unit mass emission of a gas compared to that expected from the same mass emission of CFC-11. The broad suggestion for this indicator reflects both consensuses on the methodology for its calculation and a broad scientific acknowledgment of the impact caused by the depletion of the ozone layer on humans (e.g. increased skin cancer cases) and plants. Substances causing ozone depletion have been listed in the Montreal Protocol on Substances that Deplete the Ozone Layer, which entered into force already on January 1, 1989.87
The photochemical oxidant formation potential, ozone depletion potential, eutrophication potential, and acidification potential were often suggested in the reviewed frameworks. Photochemical ozone formation directly or indirectly impacts human health via the generation of ozone at the ground level. To measure photochemical ozone formation impacts, photochemical ozone creation potential was suggested by two frameworks in the early stage of development. The LOTOS-EUROS88 model is the most common method behind this indicator. Using this model, the photochemical ozone creation potential is determined by comparing the rate at which a unit mass of chemical reacts with a hydroxyl radical (OH˙) to the rate at which a unit mass of ethylene reacts with OH.
The reviewed frameworks have frequently reported indicators for eutrophication and acidification impacts. These indicators are often considered when comparing bio-based and petrochemical alternatives.70,89–93 In fact, eutrophication and acidification impacts are usually higher for bio-based alternatives than petrochemical ones. Eutrophication is due to the release of nutrients to soil or freshwater due to fuel combustion and fertilizers in agriculture. In aquatic compartments, such nutrient excess causes the growth of algae or other plants, limiting the development of the original ecosystem. Models for the calculation of eutrophication indicators can provide a single value with no distinction per compartment94 or a separate result for freshwater and marine compartments95 and terrestrial compartment.96
Available models for calculating acidification potentials usually refer to terrestrial acidification due to atmospheric deposition of acidifying compounds.96 Terrestrial acidification is a global threat to plant diversity.97 The most significant source of acidification is fuel combustion processes, especially for fuels with a high sulphur content as those used e.g. in tractors.
Regarding resources, the reviewed frameworks have often considered resource use/depletion – fossil (MJ), water depletion potential and land use indicators. For the depletion of fossil resources, the scarcity/resource depletion model in ref. 98 is implemented in most LCA midpoint methods. The same model can also provide an indicator for the depletion of metal and mineral resources suggested in 14 of the reviewed frameworks.
The method underpinning the water depletion potential indicator suggested by most frameworks is the Swiss Ecological Scarcity Method.99 However, the AWARE model100 has emerged more recently and it is currently recommended by the European Commission.101 This model provides an indicator with the same unit as the Swiss Ecological Scarcity Method but with significantly different modelling of the characterization factors.
The indicators related to land use mentioned by the reviewed frameworks are based on various methods and models such as the Swiss Ecological Scarcity Method,99 Ecoindicator 99,102 the Soil Organic Matter model103 and the LANCA model.104 Also for land indicators, different models are not directly comparable even if the indicator might have the same unit since the characterization modelling often focuses on a different land-use aspect and covers different land types.
As also remarked as relevant by the EU CSS, various frameworks propose life cycle midpoint indicators for aspects typically considered in safety/risk assessments like ecotoxicity and human toxicity. The suggested indicators are based on various methods and models: the ReCiPe 2016 impact method,95 the USEtox model,105 the CML 2001 method94 and EDIP97.106 Indicators for toxicity aspects based on a different method/model provide significantly different results even when expressed for the same unit e.g. some do not consider certain compartments or do not consider acute (i.e. short-term) toxic effects in the ecotoxicity category.
Despite some LCA indicators for toxicity aspects being suggested by various frameworks, this should not lead to the thinking that LCA can replace risk assessments to evaluate whether a process is safe.35 In fact, there is no direct equivalency between LCA toxicity-related midpoint impacts and outcomes from risk assessments. For example, LCA does not generally consider the direct exposure pathways from a product but through exposure in environmental media.62 However, there are some attempts to bridge the gap between LCA and risk assessment.107–110
Another indicator proposed by various frameworks is particulate matter expressed in relative human health damage compared to fine particulate matter (PM 2.5 eq.) based on the model described by Rabl et al.111
Some of the frameworks suggest impacts at the endpoint level, based on damage-oriented modelling regarding three protection areas i.e., human health, ecosystems and resources, via integrated assessments (Table 3). These indicators are mainly based on two impact assessment methods i.e., ReCiPe 2016 (or earlier 2008 version) and Ecoindicator 99 (considered a precursor of the current ReCiPe 2016). The ReCiPe 2016 human health endpoint indicator (as well as Ecoindicator 99) addresses the damage to human health caused by respiratory and carcinogenic effects from organic and inorganic substances, human health issues caused by ionizing radiation, and climate change and ozone depletion. The ReCiPe 2016 ecosystem quality endpoint indicator (as well as Ecoindicator 99) addresses the damage to the ecosystem quality caused by ecotoxicological effects, land-use-related impacts, acidification and eutrophication.
Two frameworks also mentioned indicators at the endpoint level for climate change.18,112 However, endpoint indicators for individual environmental aspects are much less commonly used in LCAs than midpoint indicators.
Some aspects are suggested at the conceptual level without suggesting a specific indicator and method in the reviewed frameworks. Ten frameworks remarked the importance of protecting biodiversity. In particular, the guides by BASF54 and the US National Research Council35 consider biodiversity conservation as one of the leading sustainability criteria. However, there is a lack of data or shared consensus on monitoring biodiversity losses via current LCA indicators.35 Nonetheless, impacts on biodiversity are quantitatively strictly related to LCA endpoint indicators for damage to the ecosystem quality mentioned in the previous section.113 Various (purely) conceptual frameworks have remarked the relevance of accounting for climate change issues, eco-toxicity, human toxicity, land use, and fossil/mineral resources.
The frameworks in the early stage of development showed a similar trend to the overall indicators adopted for the evaluation of the environmental dimension. This could be due to the fact that most of the indicators refer to midpoint impact categories of LCA. LCA can be performed also in the early stage as well as the estimation of the impact categories, being aware of higher uncertainty linked to the data availability and quality (see section 3.5 for further details). The slightly higher use of indicators on pressure was observed highlighting the higher availability of water and air emission information.
Stakeholder category | Number of frameworks adopted | Early stage application | Stakeholder category | Number of frameworks adopted | Early stage application |
---|---|---|---|---|---|
Aspect/indicator | Aspect/indicator | ||||
Social | 189 | 20 | Risk of conflicts | 1 | |
Consumer | 8 | 3 | Value added | 1 | |
Brand communication | 1 | 1 | Youth illiteracy | 1 | |
Consumer acceptance | 1 | 1 | Value chain actors | 13 | 4 |
Consumer health & safety | 1 | Fair competition | 1 | ||
Content of natural substances (%) | 2 | 1 | Promoting social responsibility | 2 | 1 |
Ethics in marketing communication | 2 | Regional materials | 2 | 1 | |
Impact on basic needs of customers | 1 | Supply chain responsibility score | 3 | 1 | |
Local community | 41 | 5 | Tracking capacity | 5 | 1 |
Access to basic needs | 2 | Workers | 110 | 8 | |
Certified environmental management system | 1 | Accident rates at the workplace | 1 | ||
Community acceptance | 2 | Age | 2 | ||
Drinking water coverage | 1 | Annual job training | 2 | ||
Embodied forest area footprints | 1 | Association and bargaining rights | 1 | ||
Embodied agricultural area footprints | 1 | Child labour | 12 | 2 | |
Extraction of material resources | 1 | Disability | 2 | ||
Human rights (conceptual) | 5 | 1 | Equal opportunities and discrimination | 9 | |
Human rights (LCA impact category) | 2 | Evidence of violations of laws and employment regulations | 1 | ||
Human satisfaction (appropriateness for culture and level of noise and vibration) | 2 | 1 | Fair salary | 9 | |
Impact on the local economy | 1 | Forced labour | 11 | 1 | |
International migrant stock | 1 | Freedom of association and collective bargaining | 7 | 1 | |
International migrant workers in the sector | 1 | Gender wage gap | 1 | ||
Level of industrial water use | 1 | Labour influence | 1 | ||
Local employment | 11 | 2 | Men in the sectoral labour force | 1 | |
Net migration rate | 1 | Noise reduction | 4 | ||
Pollution level of the country | 1 | Part-time work | 4 | ||
Public welfare and safety | 2 | Presence of sufficient safety measures | 1 | ||
Respect to indigenous rights | 1 | Rate of injuries | 3 | ||
Respect to the living conditions | 1 | 1 | Respect to the national standards for security and social responsibility | 6 | 1 |
Sanitation coverage | 1 | Sexual harassment | 4 | ||
Unemployment rate | 1 | Social security and expenditure | 1 | ||
Society | 17 | Time of exposure | 1 | ||
Active involvement of enterprises in corruption and bribery | 1 | Trade unionism | 1 | ||
Contribution to economic development | 1 | Trafficking in person | 1 | ||
Corruption prevention initiatives | 4 | Weekly hours of work per employee | 1 | ||
Health expenditure | 1 | Women in the sectoral labour force | 1 | ||
Illiteracy rate | 1 | Workers affected by natural disasters | 1 | ||
Life expectancy at birth | 3 | Workers’ health & safety | 11 | 3 | |
Poverty alleviation | 1 | Working conditions (LCA impact category) | 1 | ||
Public expenditure on education | 1 | Working hours (e.g. maximum)/work–life balance | 9 | ||
Public sector corruption | 1 |
The 31 frameworks including the social dimension have often flagged aspects to be considered without proposing an indicator quantifying them based on a specific method. One of the reviewed studies remarked the lack of quantitative social assessments in common alternative assessment frameworks.114
As shown in Fig. 3, social impacts related to workers have the highest coverage in the revised framework, as 59% of total mentions of social aspects in the reviewed frameworks concerns the category “workers”. The other stakeholder categories (local communities, value chain actors and society) have a lower coverage and the stakeholder category “children” (included in the last update of the UNEP Guidelines in 2020) is not represented at all. The higher coverage of aspects related to workers can be explained by the higher data availability for work-related aspects, which usually are also easier to measure through quantitative indicators. Impacts on local communities, while being very relevant when assessing sustainability of product alternatives, are usually more difficult to assess due to the need for site-specific data. Impacts on society and value chain actors can also be challenging to assess given that in some cases the impact pathway is less defined. For what concerns impacts on consumers, they are to a large extent covered under the safety assessment.
Fig. 3 Comparison of the coverage of social aspects in the reviewed frameworks and in the UNEP Guidelines on S-LCA32, considering the various stakeholders’ categories. In the case of S-LCA, shares refer to the total number of impact subcategories recommended in the UNEP Guidelines. For the reviewed frameworks, shares refer to the mentions of social aspects concerning the six stakeholder categories. A detailed list of aspects is available in the ESI.† |
Fig. 3 also lists additional social aspects found in the literature review that are not explicitly or completely addressed in the UNEP Guidelines.
The social aspects that are included in the highest number of frameworks are child labour,12 forced labour,11 workers health and safety11 (in the stakeholder category “workers”) and local employment11 (under the stakeholder category “local community”). For the stakeholder category “value chain actors” the tracking capacity is included in 5 frameworks, while under the category “society” the aspect included the most is corruption prevention initiatives (4 frameworks), while the other two frameworks include other corruption-related aspects. For the “consumers” category the aspects ethics in marketing communication and content of natural substances are both included in two frameworks.
For what concerns positive impacts, which should capture the potential value for society or other stakeholders arising from a production and/or consumption activity, only the aspects local employment and contribution to economic development are included in the reviewed frameworks. While positive impact assessment is poorly implemented in practice (also due to the multiple conceptual definitions that can be adopted), there is clear interest in including this perspective in the sustainability assessment.115
Table 4 also shows that social dimension is seldom included in the sustainability assessment in the early stage of development. Among the indicators adopted in the early stage, worker-related aspects are the most assessed by few authors either from academia53,116–118 and international organization.119
Aspect | Number of frameworks adopted | Early stage application |
---|---|---|
Indicator | ||
Economic | 143 | 63 |
External cost | 24 | 4 |
Externality cost [€] | 3 | |
Life cycle cost [€] | 17 | 3 |
Waste (incl. emissions)/recycling treatment cost | 4 | 1 |
Internal cost | 59 | 27 |
(Total) production cost [€] | 26 | 15 |
Cost of maintenance/repairs | 3 | 1 |
Product cost | 6 | 1 |
Purchase cost | 20 | 7 |
SSbD implementation costs | 1 | 1 |
Total Annual Cost (TAC) | 3 | 2 |
Profitability | 25 | 14 |
Added value [€] | 1 | |
Financial profit [€] | 4 | 2 |
Minimum selling price [€] | 4 | 2 |
Net present value [€] | 4 | 4 |
Normalised added value [-] | 1 | |
Payback period [years] | 1 | 1 |
Profitability (conceptual) | 8 | 3 |
Total capital investment | 1 | 1 |
Yield | 1 | 1 |
Value chain actors | 5 | |
Product performance | 1 | |
Stakeholder requirements | 1 | |
Transparency and information | 1 | |
Value chain collaboration | 1 | |
Willingness to pay | 1 | |
Other | 38 | 23 |
Additional income (incentives, flexibility, and additional area) | 1 | |
Affordability | 1 | |
Breakeven point | 1 | 1 |
Comfort of occupants | 1 | |
Customer acceptance and satisfaction | 1 | |
Discounted cash flow rate of return | 1 | 1 |
Feedstock price | 2 | 2 |
Flash point | 1 | 1 |
Initial and maintenance budget | 1 | |
Innovation potential (by number of publications) | 1 | 1 |
Market acceptance | 2 | |
Non-construction cost (tax, financial cost) | 4 | 1 |
Performance uncertainty (material never used in a context) | 1 | 1 |
Point of explosion | 1 | 1 |
Predictability | 1 | |
Process cost | 2 | 2 |
Projected price | 1 | 1 |
Reaction and resistance to fire | 1 | 1 |
Scalability | 1 | |
Waste management cost | 7 | 6 |
Total Capital Cost (TCC) | 3 | 2 |
Total Production Cost (TPC) | 3 | 2 |
As shown in Table 5 indicators under the economic dimension are related to external cost, internal costs, profitability, value chain actors and others. The indicators related to internal costs are included in the highest number of frameworks59 and in particular the total production cost is mentioned in 26 frameworks.
Profitability was remarked as a relevant concept in various frameworks. Four frameworks include financial profit as a quantitative indicator to measure it, while four frameworks proposed the indicator net present value. In 8 frameworks profitability was included without specifying a quantitative indicator to measure it.
The life cycle cost was recommended in 17 reviewed frameworks. In most of these frameworks, the life cycle cost calculation was combined with environmental LCA. Several frameworks, especially from scientific articles, mentioned accounting for the externality cost and the cost of waste generated. Potentially, methods for calculating life cycle costs could include externality costs caused by life cycle environmental impacts and land eco-remediation. Analogously, social LCA impacts such as worker safety and health protection could be included in life cycle cost methods.
As observed in a critical evaluation of economic approaches performed in the EU project Orienting,§§120 a variety of Life Cycle Costing (LCC) methods have been proposed in the literature. The three main types of LCC include: conventional LCC, environmental LCC, and social LCC. This methodology, however, still lacks a general standard that provides guidelines for its use/application.121
Table 6 shows the economic indicators that have been detected in the frameworks revised in this study and their comparison with those reported in two reviews of sustainability assessment methodologies.122,123
Economic indicator | Alejandrino et al. 2021122 | Visentin et al. 2020123 | Revised frameworks | |
---|---|---|---|---|
External cost | Externality cost | ✓ | ✓ | |
Life cycle cost | ✓ | ✓ | ||
Waste (incl. emissions)/recycling treatment cost | ✓ | |||
Internal cost | (Total) production cost | ✓ | ✓ | ✓ |
Cost of maintenance/repairs | ✓ | ✓ | ||
Product cost | ✓ | |||
Purchase cost | ✓ | |||
SSbD implementation costs | ✓ | |||
Total Annual Cost (TAC) | ✓ | |||
Electricity cost | ✓ | |||
Cost of capital | ✓ | |||
Raw material cost | ||||
Labour cost | ✓ | |||
Profitability | Added value | ✓ | ✓ | |
Financial profit | ✓ | ✓ | ||
Minimum selling price | ✓ | |||
Net present value | ✓ | ✓ | ✓ | |
Normalised added value | ✓ | |||
Payback period | ✓ | ✓ | ||
Profitability | ✓ | ✓ | ||
Total capital investment | ✓ | ✓ | ||
Yield | ✓ | |||
Internal rate return | ✓ | |||
Revenues | ||||
Value chain actors | Product performance | ✓ | ||
Stakeholder requirements | ✓ | |||
Transparency and information | ✓ | |||
Value chain collaboration | ✓ | |||
Willingness to pay | ✓ | |||
Other | Additional income (incentives, flexibility, and additional area) | ✓ | ||
Affordability | ✓ | |||
Breakeven point | ✓ | |||
Comfort of occupants | ✓ | |||
Customer acceptance and satisfaction | ✓ | |||
Discounted cash flow rate of return | ✓ | |||
Feedstock price | ✓ | |||
Flash point | ✓ | |||
Initial and maintenance budget | ✓ | |||
Innovation potential (by number of publications) | ✓ | |||
Market acceptance | ✓ | |||
Non-construction cost (tax, financial cost) | ✓ | |||
Performance uncertainty (material never used in a context) | ✓ | |||
Point of explosion | ✓ | |||
Predictability | ✓ | |||
Price | ✓ | |||
Process cost | ✓ | |||
Projected price | ✓ | |||
Reaction and resistance to fire | ✓ | |||
Scalability | ✓ | |||
Waste management cost | ✓ | |||
Total Capital Cost (TCC) | ✓ | |||
Total Production Cost (TPC) | ✓ | |||
Economic impact score | ✓ | |||
Financial incentives | ✓ | |||
Risk | ✓ | |||
GDP/contribution to GDP | ✓ | |||
Investment | ✓ |
The comparison shows that a variety of indicators can be applied, depending on the scope and the perspective of the economic analysis. The assessment of externalities is still poorly implemented, while profitability indicators are included in the three reviews under considerations, showing that, at this point, the methodology is mainly applied to assess company-related financial performance, rather than actual sustainability impacts.
The economic dimension is also addressed in frameworks regarding the early stage of development. In total, 34 frameworks include aspects, mostly on profitability and internal costs, mostly by academia. Smith et al. are the only ones from an international agency introducing the GREENSCOPE indicators for the design including also indicators for the economic dimension.81
The frameworks reviewed often pointed out that a cradle-to-grave comparison of the final application (product or service) is necessary to evaluate chemicals’ safety and sustainability compared to the alternatives. However, the reviewed frameworks rarely provided clear recommendations on when a cradle-to-gate comparison of chemicals is considered enough and when a cradle-to-grave LCA evaluation becomes necessary.
To optimize the time needed to conduct an LCA, several scientific articles77,126,128,129 presented various easy-to-use LCA-based tools allowing preliminary environmental profiling, especially for the early stage of development. Examples of them include: (i) the FLASC tool calculates preliminary cradle-to-gate impacts for eight impact categories for a wide range of materials commonly used in drug manufacture;77 (ii) the Q-SA√ESS (Quick Sustainability Assessment via Experimental Solvent Selection) methodology calculates six cradle-to-grave sustainability metrics for the three “most sustainable solvents” for a specific process;130 and, (iii) the US EPA (United States Environmental Protection Agency) created a method rapidly generating life cycle inventories from publicly available databases by allocating the emissions from facilities related to the production of the chemical of interest.131
Other leading streamlined LCA tools are the ecosolvent tool132 for solvents, the LICARA NanoSCAN tool133 for nanomaterials and other models proposed by recent literature for application to a broad range of chemicals (e.g. ref. 126). Tools for streamlined LCAs can provide valuable decision support for chemicals in their early stage of development when data availability is very limited. However, the results generated using such tools have high uncertainty especially due to low technological, geographical and temporal representativeness. Hence, robust evaluations can be generated only via full LCAs.
Alternatively, previous studies128,134 proposed the use of physicochemical properties to predict the life cycle environmental impact in the early stage of development. Their approach assumes that there is a link between those properties and the environmental performance of the chemical production process being developed and assessed. Finally, Pizzol et al.135 recently proposed and tested a tiered approach with qualitative assessment for safety, environmental, and social dimensions in the early stage of development.
While several environmental LCAs of nanotechnologies have already been published,136,137 various studies acknowledged the challenges of conducting LCAs of nanomaterials due to their complexity and dynamic behavior during the life cycle.138–140 However, an effort is currently ongoing to fill this gap. Such an effort is ongoing also for LCAs covering social and economic aspects.141 In particular, LCA guidelines for manufactured nanomaterials were released in 2018.142
The environmental assessment of chemicals has been evolving and moving from typical green chemistry mass-based metrics to a life-cycle perspective, as this was identified as indispensable to verify actual environmental benefits.17,130,143
The preliminary step in frameworks assessing safety and sustainability of alternative chemicals is the identification of alternatives and its technical performance. The technical performance in fulfilling the function of the candidate alternative and of the alternatives in providing such functions is established via techno-feasibility assessments.18,35,144,145 New or alternative chemicals should be compared based on equal functional performances using “substitution factors” and for LCA using a “functional unit”. Nevertheless, a calculation procedure for substitution factors for a specific function and/or a structured method to detect respective alternatives was rarely reported.
Then, the safety performance of the alternatives scrutinizing physicochemical properties and applying risk assessment is evaluated. If safety is part of the framework, the evaluation of environmental, social, and economic aspects is conducted only for chemicals passing the safety assessment.
In the case of social impact assessment, compensation between positive and negative impacts should be avoided. Moreover, when assessing positive impacts, great caution must be taken with the inclusion of product utilities and when comparing the positives for one stakeholder group with the negatives for another. Indeed, as observed by Croes et al.146 an imprudent inclusion of positive impacts might led to whitewashing practices and loss of credibility of the assessment.
The vast majority of frameworks provide a separate outcome for each aspect considered or at least per dimension (safety, environmental, social, and economic). For example, a chemical can have the outcome “recommended” in the environmental dimension but “problematic” in the safety dimension or vice versa.62 The decision is then left to the user of the outcome, leaving an appropriate degree of freedom on the final decision. In particular, if safety is part of the assessment, an aggregated score over multiple sustainability dimensions is not recommended to avoid compensation between different impacts. To facilitate decision making, the impact profiles of the alternatives can also be presented at the highest aggregated level with single scores per dimension.18,62
The score is often translated in colour coding based on a percentage performance indicator, e.g. 0% representing the no sustainability (the alternative performs the worst in that aspect) and 100% representing the highest sustainability (the alternative performs the best in that aspect), and this is particularly the case of frameworks developed for the early stage.81 Pfizer was one of the first companies to use color-coding to categorize solvents (green = preferred, amber = usable and red = undesirable).147 A similar coding system has also been proposed by other companies e.g. Sanofi, Astra Zeneca and GSK77,148,149 and environmental agencies e.g. the German Environment Agency.150
In most cases, color-coding is applied to the outcome of each criterion to evaluate safety, health, and environmental aspects.73,148,150 For example, each chemical can get a score between 1 and 10 for each criterion, which is then translated into the 3-color code (e.g. green, yellow, and red).62,73,130,148,149,151 Except for green, the meaning of the other colours can be slightly different, e.g. red can mean undesirable147 or substitution requested.148,150 Some guides use brown to catalogue banned chemicals (e.g. ref. 148). White colour is often used for data unavailability that does not allow the assessment for a certain criterion.148–150 Sometimes also orange is included as a colour to indicate a chemical that should be substituted but does not have to if it is still compliant with current regulation.73
Other approaches to support decision making suggested in the literature include the use of Multi-Objective Optimization (MOO) techniques or Multi-Criteria Decision Analysis (MCDA). MOO frameworks for alternative assessment are normally implemented in computer-aided molecular design tools using for example a Fuzzy Analytic Hierarchy Process (FAHP) weighting approach.63,152 This means that the decision-making is structured as a hierarchy where the primary goal of the design e.g. safety comes before other criteria and sub-criteria, giving priority to the objects of the decision problem that must be fulfilled.63,152 Regarding the design, simple hotspot analysis is also conducted to guide further development of the design of the new chemical or material.153
MCDA, which allows simultaneous comparison of multiple and often conflicting aspects, has also been highlighted as a key instrument for sustainability assessment, as discussed in major works and reviews.29,30,154 Two commonly used MCDA methods are the multi-attribute utility theory and outranking.61 Although MCDA methods may be useful in providing decision makers with a common baseline to understand the performance of alternatives and the trade-offs they present, they may be significantly resource intensive. MCDA in the early stage of development have been found to be recently explored by few authors. Garas et al.117 adopted a Sustainable Decision Support System (SDSS) scoring system that integrates LCA and MCDA; García-Velásquez C.155 used the Pareto frontiers to guide decision in the plastic sector. Finally, Manjunatheshwara and Vinodh adopted a grey method for the decision specifically for materials selection at the design phase with uncertain conditions.156
Some of the frameworks propose ways to deal with data gaps, reflecting this in the evaluation. Malloy et al.61 applied an MCDA framework to assess the impact of data gaps on alternative assessment using multi-attribute utility theory and outranking other tools, penalising aspects with missing data by applying a lower score (GreenScreen full assessment, SciVera, GreenSuite); other tools, usually list-based tools, consider missing data as undetermined (GreenScreen List Translator) or indifferent on the final score (GreenWERCS).157
GreenSuite's procedure uses five criteria to differentiate the cause of the missing data to score the aspect as more hazardous.158
The GreenScreen® tool (and by extension, the IC2159 and Rossi44 frameworks) propose a system based on the preliminary score to assess the level of the material analysed, in which data gap analysis is applied to determine if the data requirements are met. If the analysis fails, the final score is lowered by one unit, otherwise the score is confirmed.45
OECD160 addresses data gaps by using two different approaches, depending on whether the data quality is limited (tier 1) or whether high quality data are used (tier 2), stating the quality of the assessment to the audience.
Regarding uncertainty assessment, a limited number of frameworks have suggested ways to perform it. NRC157 suggests a list of good practices to deal with uncertainties that include the selection of alternatives with only known aspects and conducting a quantitative analysis, pointing out that when uncertainty is large enough to overwhelm any relative differences between alternatives, it becomes impossible to determine any better alternative. Safe Consumer Products161 provide a stepwise approach to carry-out uncertainty assessment and data gaps.
Although the assessment of chemicals or materials in the early stage of development is quite uncertain due to the lack or quality of data, only 9 authors focused on sensitivity analyses or uncertainty, most of them regarding construction and solvent sectors. Among available options, Posada et al. performed a Monte Carlo simulation to identify the variability on the input data, and similarly, Zapata Boada et al. analysed the influence of parameters affecting economic and environmental performances by sensitivity analysis. In addition, the Triangular Fuzzy Number and Fuzzy Topsis,152,162 the VEGA toolbox163 and the IDEMAT 2001 database164 have been used to evaluate variability and uncertainty. Uncertainty assessment is key for early stage assessments as it provides the decision maker the spectrum of possibilities enabling a more informed decision making. At a minimum, sensitivity analysis should be conducted on key parameters in the system to study the robustness of results and their sensitivity to uncertainty factors. This will determine whether data collection and quality need to be improved and enhance the interpretation of results.
While some of the reviewed frameworks are conceptual, other frameworks provide a detailed guideline to support the selection of safer and more sustainable chemicals. Most reviewed frameworks pointed out that the criteria regarding safety and sustainability of alternatives should be based on equal functional performance. However, they lack providing a calculation procedure of substitution factors for a specific function and a structured method to detect respective alternatives.
A major focus was on scrutinizing sustainability aspects and indicators and respective calculation methods as well as the decision procedures proposed by the frameworks. The intent was to understand the current state of art and gaps to reach a better-informed decision-making process for designing or selecting safe and sustainable chemicals. This review highlighted that there is no uniform and comprehensive set of indicators for examining the sustainability of a chemical within proposals of frameworks from academia, governments, NGOs, or industry, especially for what concerns socio-economic aspects. This fact could negatively impact the roadmaps of chemicals since they might be sustainable according to one framework but not another.
In this sense, LCA can be of use as it covers multiple environmental impacts. In fact, there was a broad consensus on the need to account for the life cycle of chemicals and on the need to use indicators based on the life cycle assessment methodology. In fact, LCA can overcome the limitation of simple mass- and energy-based metrics that do not capture actual shifts in environmental burdens by selecting an alternative instead of another. LCA has been gaining prominence in sustainability assessment nonetheless there are limitations that need to be addressed to ensure robust assessments. In particular, guidance is needed for LCA modelling of technologies at a low technology readiness level and for which the data gaps and uncertainty are even more predominant. While the S-LCA does not have the same level of maturity as the environmental LCA, this methodology underpins internationally agreed guidelines that can be taken as a reference, especially for what concerns the list of social aspects to be selected for the assessment, and as general guidance for the social assessment. The LCC methodology is the most heterogeneous for what concerns the methodological approach but also from a conceptual point of view (which kind of impacts should be assessed, area to be protected, etc.).
Increasing chemicals’ circularity is also acknowledged by the EU CSS as a way to contribute to reducing chemical pollution in wastewaters. However, mass-based/circularity metrics in the reviewed frameworks do not account for the effect of multiple cycles in environmental assessment as well as hazard and risk assessments. Therefore, as also remarked by the EU CSS, there is a need to develop methodologies for chemical risk assessment that take into account the whole life cycle and the effect of increased circularity.
To have the “paradigm shift” towards safe and sustainable chemicals, the industry and sustainability/LCA community need to respond to the challenges resulting from this review. Numerous organizations already have many initiatives, but these are carried out mainly independently. With a lack of coordination, it is difficult to guarantee a harmonized selection of suitable sustainability indicators to be integrated into future frameworks. This review shows that there is no uniform set of indicators within proposals of frameworks from academia, governments, NGOs, or industry for evaluating the sustainability of chemicals. If different indicators are implemented in the various frameworks developed in parallel for the same context, they can negatively impact the product roadmaps that often take years for development.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3gc04598f |
‡ Current affiliation: Syensqo Lyon Research and Innovation Center, 85 Avenue des Freres Perret 69190 Saint-Fons, France. |
§ TITLE-ABS-KEY ((“alternatives assessment” OR “chemicals alternative assessments” OR “alternatives analysis” OR “substitution assessment” OR “chemicals assessment” OR “solvent selection” OR “solvents selection” OR “solvent design” OR “safe and sustainable” OR “social LCA” OR “life cycle costing” OR “life cycle cost”) AND (“chemical” OR “chemicals” OR “solvent” OR “solvents”) AND (“framework” OR “frameworks” OR “guide” OR “guides” OR “methodology” OR “methodologies” OR “tool” OR “tools”)). |
¶ TITLE-ABS-KEY (multicriteria OR multi-criteria OR “multiple criteria” OR mcda OR mcdm OR multiattribute) AND (chemical OR material OR substance) AND (safe* or sustainab*). |
|| Chemicals are substances and mixtures as defined in Registration, Evaluation, Authorization and Restriction of Chemicals (REACH) and Classification, Labelling and Packaging (CLP) legislations. |
** Materials denote either substances or mixtures which may or may not yet fulfil the definition of an article under REACH and may be of natural or synthetic origin.24 |
†† Products are goods supplied for distribution, consumption or use on the Community market whether in return for payment or free of charge (EU Ecolabel). Materials denote either substances or mixtures which may or may not yet fulfil the definition of an article under REACH and may be of natural or synthetic origin.24 |
‡‡ Annex 10 of the guide proposed by the German Environment Agency34 provides a comprehensive overview of the energy consumption of chemicals and materials. In this guide, “green chemicals” are chemicals consuming less than 10 MJ kg−1 during production, “yellow chemicals” between 10 and 100 MJ kg−1 and “red chemicals” more than 100 MJ kg−1. |
§§ Operational Life Cycle Sustainability Assessment Methodology Supporting Decisions Towards a Circular Economy, grant agreement no 958231. |
This journal is © The Royal Society of Chemistry 2024 |