Chemicals Alternatives Assessment (CAA): Tools for Selecting Less Hazardous Chemicals

Chemicals alternatives assessment (CAA) is a form of alternatives assessment that focuses on finding alternative chemicals, materials or product designs to substitute for the use of hazardous chemicals in products. Chemical hazard assessment (CHA) or comparative CHA is a method for comparing chemicals based on their inherent hazard properties. CAA is inclusive of CHA. CHA/CAA provides users with hazard‐based information to make informed decisions when selecting less hazardous chemical alternatives. The approach is used to assess a chemical's impact on human health and the environment. Hazard can be defined as the way in which a chemical, object or situation may cause harm. The degree of a chemical's capacity to harm depends on its intrinsic properties, such as its capacity to interfere with normal biological processes and its capacity to burn, explode or corrode (e.g., non‐life‐threatening allergic skin reaction to nickel jewelry, lethal egg‐shell thinning in avian species attributed to exposure to the organochlorine pesticide DDT).¹  The goal of a CAA is to find a science‐based solution that identifies and completely characterizes chemical hazards, promoting the selection of less hazardous chemical ingredients, in addition avoiding unintended consequences of switching to a poorly characterized chemical substitute.

Hazard assessment is a systematic process of assessing and classifying hazards across an entire spectrum of endpoints and levels of severity It involves a characterization of the nature and strength of the evidence of causation.²  A comparative CHA is a type of hazard assessment that evaluates hazards from two or more agents, with the intent to guide decision‐making toward the use of the least hazardous options via a process of informed substitution, as illustrated in Figure 1.1.

In practical terms, comparative CHA is a term that describes the practice of assessing hazards for specific )s (such as chemicals, materials, products or technologies) and then comparing these hazards following a structured approach. Ideally, CHA minimizes subjectivity in hazard classification since a structured approach is used to assign hazards. Over the past 10 years, the number of comparative hazard tools has continued to increase. The Toxics Use Reduction Institute (TURI) at the University of Massachusetts presented a collection of over 100 tools for comparing hazard characteristics of different chemicals.³  CAAs have numerous applications, including the following:

• Enabling the prioritization of chemicals for reduction or phase‐out:

  • from any phase in product life‐cycle (e.g., manufacturing, product design);

  • from the whole supply chain.

• Assisting in the selection of alternatives for the following:

  • banned or restricted chemicals or materials;

  • chemicals that are perceived as hazardous by the public;

  • identifying and classifying Restricted Use Materials (RUMs);

  • developing environmentally preferred products.

Some CHA methods, such as GreenScreen™, focus solely on individual chemicals or materials and their subsequent health or environmental impacts while other CAA methods such as Cradle to Cradle® incorporate CHA in addition to certain life‐cycle‐based considerations such as energy use, water quality and efficiency, social responsibility and potential for material reuse.

To date, various CAA partnerships have brought together environmental agencies, such as the US EPA, industry organizations such as the Phosphorus, Inorganic and Nitrogen Flame Retardants Association (pinfa), academia (such as the University of Massachusetts at Lowell) and non‐governmental organizations such as the United States Green Chemistry and Commerce Council (GC3) and Europe's ChemSec to evaluate environmental and health impacts of potential alternatives to problematic chemicals and chemical classes, such as phthalate esters (ubiquitously used in flexible plastics), flame retardants in furniture and printed circuit boards and nonylphenol ethoxylate surfactants (which are commonly used in laundry detergents and are exceedingly toxic to aquatic organisms). Such partnerships demonstrate that CAA can be employed to benefit both producers and users of the chemical to the improvement of ecological and human health.

The purpose of this collection of chapters in the Issues in Environmental Science and Technology series is to describe and exemplify several existing CAA methodologies currently being used in North America, Europe and China and to make suggestions on how to improve the overall CAA process, fostering the greater adoption of CAA around the world. This introductory chapter identifies a number of common themes among CAA methods and provides a broad overview of such methods.

CAA paradigms share a number of similarities that are implemented as part of a CAA and all have the common goal of identifying less hazardous chemicals. CAAs use standardized procedures to assess whether alternatives have the potential for an improved health and environmental profile. CAAs assess whether the adoption of an alternative chemical is likely to result in lasting environmental or public health improvement. Ideally, a CAA will also address whether chemical alternatives are commercially available, perform well and are cost‐effective.

As a first step towards characterizing the human health and/or environmental hazards of a chemical, a CHA is performed. The practitioner assesses hazards for each chemical alternative across a range of health effects and environmental endpoints. Such endpoints generally include the following: acute and repeat dose toxicity, endocrine activity, carcinogenicity and mutagenicity, reproductive and developmental toxicity, neurotoxicity, respiratory and dermal sensitization, skin and eye irritation, acute and chronic aquatic toxicity, terrestrial toxicity and persistence and bioaccumulation. When measured data are not available or adequate for an endpoint, a hazard concern level can be assigned based on quantitative structure–activity relationships (QSARs or SARs) and expert judgment. This practice ensures that all endpoints are considered as part of the hazard assessment and that alternatives are evaluated based on a complete understanding of their potential human health and environmental hazards. A level of confidence associated with studies is often assigned.

Sources of information to evaluate and characterize human health and environmental hazards in a CAA include one of more of the following:

• Publicly available experimental data obtained from a literature review

  • Sources of such toxicological and environmental fate and effects data include online databases indexing scientific literature, such as:

    • ChemIDplus: http://www.cleanproduction.org/library/greenscreen‐translator‐benchmark1‐possible%20benchmark1.pdf

    • EPA High Production Volume Information System (HPVIS): http://www.epa.gov/hpvis/index.html

    • UNEP OECD (Organisation for Economic Co‐operation and Development) Screening Information Datasets (SIDS): http://www.chem.unep.ch/irptc/sids/OECDSIDS/sidspub.html

    • European Chemical Substances Information System IUCLID Chemical Data Sheets (ESIS): http://esis.jrc.ec.europa.eu/index.php?PGM=dat

    • United States National Toxicology Program (NTP): http://ntp.niehs.nih.gov/

    • International Agency for the Research on Cancer (IARC): http://www.inchem.org/pages/iarc.html

    • Human and Environmental Risk Assessment on ingredients of household cleaning products (HERA): http://www.heraproject.com/RiskAssessment.cfm

    • European Chemicals Agency (ECHA): http://echa.europa.eu/

    • ExPub (Expert Publishing): http://www.expub.com

  • Experimental data that are not publicly available (such as industry‐ or trade association‐sponsored studies)

  • SAR‐based estimations from predictive methods such as US EPA models (e.g., EpiSuite, Ecosar), European Union (e.g., VEGA, ToxTree) or OECD (e.g., OECD Toolbox), Derek, Topkat, among other predictive software algorithms.

Once the literature search has been performed, relevant studies have been retrieved and data collected, the second step of a CHA generally entails assigning hazard scores for the criteria evaluated. For example, a GreenScreen™ will assign hazard scores of low, moderate, high (and, for some endpoints, very low or very high) for 18 health and environmental fate and toxicity endpoints, as illustrated for an example chemical in Figure 1.2.⁴  Criteria for assigning hazard scores in a CAA are often based on the Global Harmonized System of Classification and Labeling of Chemicals (GHS) criteria,¹  in addition to criteria from other authoritative lists. As an example, a complete version of GreenScreen™'s hazard criteria for each endpoint can be found at the CPA's website.⁴ 

For several endpoints, such as acute mammalian toxicity, systemic toxicity, acute and chronic aquatic toxicity, persistence and bioaccumulation, hazard scores are often assigned based on specific dose thresholds and/or ranges. For example, the US EPA's Design for the Environment (DfE) Alternatives Assessment Criteria for Hazard Evaluation will assign a chemical a hazard score of Low for acute mammalian toxicity based on an oral LD₅₀ of 2000mg kg⁻¹ or greater and a hazard score of Moderate for persistence based on a half‐life in water that falls between 16 and 60 days.

For other endpoints, such as carcinogenicity, mutagenicity and reproductive and developmental toxicity, professional judgment and a clear understanding of the available data are necessary to draw a conclusion and assign a hazard classification. If a 2 year carcinogenicity study performed by the US National Toxicology Program (NTP) shows a statistically significant increase of hepatic tumors in rats, US EPA DfE and GreenScreen™ alternatives assessment criteria would assign a hazard score of High for carcinogenicity. In contrast, if a reproductive toxicity study reports a Lowest Observed Adverse Effect Level (LOAEL) of 250mg kg‐bw⁻¹ d⁻¹ in mice based on an effect such as decreased weight gain, US EPA and GreenScreen™ alternatives assessment criteria may or may not consider this effect relevant for purposes of assigning a hazard for reproductive toxicity depending on the chemical's mechanism of action. Ultimately, the hazard classification may come down to the professional opinion of the scientist performing the CAA. It is imperative to the integrity of a CAA that all hazard scores are based on sound scientific knowledge and can be properly supported and defended, if necessary.

Points to consider when determining the validity of available data include whether a study was performed following Good Laboratory Practices (GLP) or whether a study was conducted following a specific test guideline (e.g., OECD test guidelines). The level of detail reported in a study is also important, as is the source of the study. Primary sources such as peer‐reviewed studies are preferred; however, high‐quality secondary sources are acceptable, particularly when supported by a Klimisch score of 1 or 2 (1=reliable without restrictions; 2=reliable with restrictions), which provides an indication as to the reliability of the actual data.⁵ 

In the case of conflicting data, weight of evidence should be used to assign the ultimate hazard score for a specific health effect/environmental endpoint. All professional judgments must be fully justified within the section of that endpoint. The justification for a final hazard score must be transparent and easily understood by all who may read the CAA.

Once the hazard assessment and classification portion of the CAA has been completed, a CAA report is written to provide contextual and supplemental information designed to aid in decision‐making and may include descriptions of manufacturing processes, use patterns and life‐cycle stages that may pose special exposure concerns. The CAA report may contain a description of the cost of use and the potential economic impacts associated with the selection of alternatives and may also contain information on alternative technologies that might result in safer chemicals, manufacturing processes and practices.

Examples of alternatives assessments can be found for numerous types and classes of chemicals, including nonylphenol ethoxylate surfactants, flame retardants in furniture and printed circuit boards, diethylhexyl phthalate and perchloroethylene.^6,7 

In comparing alternatives, it is important to consider life‐cycle impacts in order to avoid shifting burdens between stages of a material's life‐cycle. Life‐cycle assessment (LCA) is a standardized and quantifiable approach to assessing life‐cycle impacts, such as ISO Standard 14040 (Environmental Management: Life Cycle Assessment‐Principles and Framework).^8,9  Production of all chemicals or substances requires the extraction of resources from the earth, including water, energy and other raw materials. Energy and other resources are used for manufacture, transportation and during the use phase. At the end of a product's useful life, the product may be released into the environment. This can be problematic if the product contains hazardous materials. As climate change becomes more of a global concern, manufacturers are beginning to assess how their processes affect the environment. Many businesses have responded by identifying ‘greener’ raw materials and using ‘greener’ processes to manufacture their products. They are able to reduce the environmental impact their actions have by utilizing LCA. There are four main phases of an LCA:

1. Defining the goals and scope of the LCA – Define and describe the product, process or activity.

2. Performing a life‐cycle inventory – Identify and quantify energy, water and materials usage and environmental releases (e.g., air emissions, solid waste disposal, waste water discharges);

3. Performing a life‐cycle impact assessment – Assess the potential human and ecological effects of energy, water and material usage and the environmental releases identified in the inventory analysis;

4. Interpretation – Evaluate the results of the inventory analysis and impact assessment to select the preferred product, process or service with a clear understanding of the uncertainty and the assumptions used to generate the results.

LCA helps to avoid shifting environmental burdens from one step in a product's life‐cycle to another. For example, when choosing between two materials, Option 1 may appear environmentally beneficial because it is more easily biodegradable. However, after conducting an LCA, it may become obvious that Option 2 will produce less of an environmental impact because it uses less energy to manufacture and its breakdown products may pose less of a risk to aquatic ecosystems. An ideal CAA paradigm will take into account life‐cycle thinking when evaluating a chemical or material and seek to reduce hazard across the life‐cycle. Commercially available LCA software packages such as GaBi and SimaPro are commonly used for LCA, with growing interest in the freeware OpenLCA, which requires inputting of life‐cycle inventory data.

Currently, there are a number of CAA paradigms in use around the world to evaluate environmental and human health impacts of potential alternatives to problematic chemicals, as presented in Table 1.1. Often, CAAs employ life‐cycle considerations to predict and understand specific phases, from development to manufacture, where industry can make changes to realize environmental and health benefits.^6,10 

As mentioned above, CAAs are inclusive of CHA and share common endpoints. CAA methods evaluate chemicals based on their measured or predicted human and ecological hazards, in addition to their environmental fate. Human health criteria in a CAA evaluate endpoints such as potential carcinogenicity, mutagenicity, reproductive and developmental toxicity, endocrine disruption, acute and chronic or repeat dose toxicity, irritation and sensitization. Acute and chronic aquatic toxicity, terrestrial toxicity, persistence and bioaccumulation are commonly evaluated to predict a chemical's environmental toxicity and fate. Finally, some CAAs (such as GreenScreen™) also evaluate a chemical's physical characteristics. such as flammability and reactivity.

Of the CAA methods listed, only the US EPA's DfE program and CPA's GreenScreen™ are fully transparent and publicly available methods of assessment. A number of the other CAA methods identified in Table 1.1 do not fully disclose all of their reasoning or resources used for establishing threshold values for hazard criteria, prioritization of hazard endpoints and life‐cycle concerns.

This section describes a number of CAA methods. Although each method uses its own set of criteria when evaluating the hazards of a chemical, the most commonly used benchmarks used when assigning a hazard for a specific endpoint mirror those developed by the US EPA, the OECD, and/or the Globally Harmonized System of Classification and Labeling of Chemicals (GHS). Tables 1.2 and 1.3 provide an overview of how each CAA paradigm evaluates human health and environmental hazard criteria, and it is important to remember that these criteria are dynamic and change over time.

The US EPA's DfE program works with industry, environmental groups and academia to reduce risk to human health and the environment.⁶  The US EPA DfE Alternative Assessments program employs a variety of design approaches to reduce the overall human health and environmental impact of a product or process. DfE's Alternatives Assessment framework is a hazard‐based assessment tool for evaluating and differentiating among chemicals based on their concern for human health and environmental hazard, in the process promoting informed substitution.^10,11  DfE's Alternatives Assessments follow six broad steps, as illustrated in Figure 1.3.¹⁰ 

As part of a DfE Alternatives Assessment, chemicals are evaluated for numerous health effect and environmental endpoints, including carcinogenicity, mutagenicity, reproductive and developmental toxicity, acute and repeat dose toxicity, toxicity to aquatic organisms and environmental fate.¹¹  For most hazard endpoints, DfE criteria define ‘High,’ ‘Moderate,’ and ‘Low’ concern. Authoritative sources [the United Nation's Globally Harmonized System (GHS) for the Classification and Labeling of Hazard Substances and US EPA programs] are the basis for these distinctions. Both experimental and modeled data are used in assigning a hazard designation of High, Moderate or Low. In the absence of experimental data, measured data from a suitable analog are preferred over estimated data. Approved modeling tools include EPI Suite, ECOSAR, OncoLogic and the Endocrine Disruptor Screening Program to predict possible hazards.

DfE CAAs do not specify a favored alternative, but do promote informed substitution when combined with cost, performance and national and international regulatory initiatives and requirements. DfE's Alternatives Assessments are used by the Office of Pollution Prevention and Toxics (OPPT) in EPA's Office of Chemical Safety and Pollution Prevention (OCSPP) to seek safer alternatives. Currently, Chemical Action Plans are the primary source for identifying chemical candidates for risk management and specifying actions. EPA proposes to further evaluate the chemicals and address risks. DfE has applied their alternatives assessment paradigm to flame retardants in furniture and printed circuit boards, along with identifying alternatives to chemicals for which there are Agency Action Plans, including nonylphenol ethoxylate surfactants, bisphenol A alternatives in thermal and paper and alternatives to the flame retardants decabromodiphenyl ether (decaBDE) and hexabromocyclododecane.¹² 

The GreenScreen™ for Safer Chemicals (version 1.2) is a quantitative chemical screening method designed to help manufacturers identify chemicals of concern and inherently less hazardous alternatives to benefit humans and the environment.¹³  The structure of the GreenScreen™ method builds from the alternatives assessment approach developed by the US EPA's DfE CAA program.¹¹  Similarly to a DfE CAA, the GreenScreen™ was developed as a comparative hazard assessment method that assists in selecting chemicals that are inherently safer for humans and the environment. Both GreenScreen™ DfEs are based on the principles of Green Chemistry¹⁴  and focus on managing chemical risk by reducing hazard rather than controlling exposure to potentially toxic chemicals. As part of a GreenScreen™ evaluation process, each ingredient or chemical is assigned a hazard concern level. Individual hazards are evaluated for 18 hazard endpoints (such as carcinogenicity, reproductive toxicity, neurotoxicity aquatic toxicity, persistence and bioaccumulation) and then a level of concern of High, Moderate or Low is assigned for each endpoint for each chemical. Two hazards, persistence and bioaccumulation, have an additional level of concern of Very High to reflect the growing international consensus in defining very persistent and very bioaccumulative (vPvB) chemicals. The threshold values for each hazard endpoint are based on those established by the EPA's DfE program and also the GHS where available. After a chemical has been evaluated against criteria for each hazard endpoint, the results are collectively scored to one of the following GreenScreen™ Benchmark scores:^16,17 

• Benchmark 1:Avoid (Chemical of High Concern).

• Benchmark 2:Use (But Search for Safer Substitutes).

• Benchmark 3:Use (But Still Opportunity for Improvement).

• Benchmark 4:Prefer (Safer Chemical).

• Benchmark ‘U’:Undetermined (Insufficient Data).

To progress from a lower GreenScreen™ Benchmark score (more hazardous) to a higher (less hazardous) GreenScreen™ Benchmark score, a chemical and its feasible and relevant transformation products must pass all criteria specified under the lower Benchmark. The criteria become increasingly more stringent for environmental and human health and safety and also for data completeness for each Benchmark. Depending on the type and number of data gaps, a chemical may either receive either a downgraded Benchmark score or be assigned Benchmark ‘U’ to indicate insufficient data. The criteria for Benchmark One align with those that governments in Europe, Canada and the USA use to characterize substances of very high concern.

Figure 1.2 identifies the hazard‐based scores for the plasticizer di(2‐ethylhexyl) terephthalate (DEHT), which result in a GreenScreen™ Benchmark score of 3: ‘Use (But Still Opportunity for Improvement).’

CPA's GreenScreen™ (version 1.0) was initially developed to assess only organic chemicals.¹⁸  Because some inorganic chemicals are recalcitrant and others are persistent as stable moieties, hazard criteria for persistence are not always relevant to inorganics. Persistence alone does not indicate that a chemical is hazardous. Chemicals that are persistent as well as bioaccumulative and toxic are of high concern, as their concentration in the environment increases over time, allowing for more opportunity to exert a toxic effect on human health or aquatic or terrestrial organisms. In 2011, GreenScreen™ (version 1.2) was expanded to address hazards from inorganics such as mineral oxides to allow for comparison of inorganic chemicals used as flame retardants.¹⁷  Under GreenScreen™ version 1.2 criteria, a persistent inorganic chemical with a Low hazard rating for human‐ and eco‐toxicity across all hazard endpoints and a Low hazard rating for bioaccumulation will not be considered problematic. Inorganic chemicals that are only persistent may reach Benchmark 4 [the best score: Prefer (Safer Chemical)]. The GreenScreen™ is not intended to address all critical elements of sustainability. It does not consider social equity or important life‐cycle impacts such as energy quantity and quality like other alternatives assessments. A GreenScreen™ assessment does not necessarily include reagents used to synthesize a chemical. Rather, the GreenScreen™ focuses on hazard assessment of chemicals from their point of generation. The GreenScreen™ is a CHA that can be used in a modular way with any other desired components of a full CAA.

The Cradle to Cradle® method (C2C) began as a proprietary product certification program that incorporates CHA and can be used for CAA. It is currently being maintained and administered by the Cradle to Cradle Products Innovation Institute (C2CPII)¹⁹  and is moving toward greater transparency. C2C uses the metabolism of Nature as a model for human industry – products or materials that cannot be metabolized by the natural world should never enter it. Rather, they should be considered as ‘technical nutrients’ and designed to flow in technical cycles. Evaluation criteria are grouped into the following categories: ‘Material Health,’ ‘Material Reutilization,’ ‘Renewable Energy and Carbon Management,’ ‘Water’ and ‘Social Fairness.’ Materials and products are certified as Basic, Bronze, Silver, Gold or Platinum, based on how the criteria are met. For the Material Health evaluation, all ingredients are ‘scored’ based on their impact on human and environmental health. Ingredients are evaluated against all common endpoints mentioned above and scored as Green (Little to No Risk), Yellow (Low to Moderate Risk), Red (High Risk) or Grey (Incomplete Data). The C2C paradigm goes further to evaluate a product based on Material Reutilization. The percentage of the product that is considered ‘Recyclable or Compostable’ is combined with the percentage of the product that is manufactured from ‘Recycled or Renewable Content’ to calculate a Nutrient Reutilization Score.

In addition to the materials used, the amount of energy and water (both quantity and quality) required for product manufacture and assembly is evaluated for certification. The ultimate goal of C2C design is to have all energy inputs come from ‘current solar income’ (i.e. geothermal, wind, biomass, hydro and photovoltaic energy sources). For Gold certification, at least 50% of purchased electricity is renewably sourced or offset with renewable energy projects, and 50% of direct onsite emissions are offset. For Silver, Gold and Platinum certification levels, the applicant must create or adopt a set of principles or guidelines to illustrate the manufacturing facility's strategies for protecting and preserving the quality and supply of water resources. An audit of the facility is required for Bronze and higher certification. For Gold certification, the applicant must demonstrate that the facility has optimized all product-related process chemicals in effluent. Finally, the organization must adopt and make public one or more statements regarding their social and ethical performance goals such as fair labor practices, corporate and personal ethics, customer service and local community outreach. As detailed in the Cradle to Cradle Certified^CM Products Standard, requirements for the Social Fairness category increase relative to the specific certification level.¹⁹ 

In 1997, the Toxics Use Reduction Institute (TURI) at the University of Massachusetts developed the Pollution Prevention Options Assessment System (P2OASys) tool to help companies determine the potential environmental, worker and public health impacts of alternative technologies aimed at reducing toxics use. The P2OASys tool is a downloadable software package that assists industry in two ways: it systematically examines the potential environmental and worker impacts of toxic use reduction options in a comprehensive manner, examining the total impacts of process changes, rather than simply those of chemical change, and compares toxic use reduction options with the company's current process based on quantitative and qualitative factors.²⁰  The user inputs both quantitative and qualitative data on a chemical's toxicity, ecological effects and physical properties. Data needed for P2OASys are available through vendors, existing databases or the Toxics Use Reduction Institute. In addition, data for up to three alternatives can be entered as a comparison with the original material. Embedded formulae in the P2OASys program can provide a numerical hazard score for each endpoint for the ingredient(s) in question. For example, if a chemical has an oral LD₅₀ of 5000mg kg⁻¹, it may be assigned a standardized hazard score of 2; the higher the score, the greater is the hazard. Endpoints include human effects, aquatic hazards, persistence/bioaccumulation, atmospheric hazard, disposal hazard, energy/resource use and exposure potential.

The Chemical Scoring and Ranking Assessment Model (SCRAM) was developed to rank or score chemicals based on persistence, bioaccumulation and toxicity.²¹  The program was initially developed for use in the Great Lakes area but is not site specific. This program consists of a simple spreadsheet system that allows the assessor to calculate an index (scores range from 1 to 5) based on the potential exposure and toxicity of chemicals. In addition, unlike other assessment tools, SCRAM addresses the uncertainty of these rankings due to a lack of data. Instead of ignoring a chemical with a large data gap, SCRAM will assign an uncertainty score that, along with its persistence, bioaccumulation and toxicity scores, is used to rank the chemical relative to others in question. In‐house libraries and on‐line databases are searched for data describing persistence, bioaccumulation and toxicity of a chemical in question. A minimum of one data point is required for bioaccumulation and is scored on the basis of bioaccumulation factors (BAF), bioconcentration factors (BCF) or the octanol/water partition coefficients (K_ow). A minimum of one data point is required for scoring a chemical in the persistence category. Persistence is scored based on half‐lives in five environmental compartments: biota, air, soil, sediment and water. Measured data take priority over estimated data; however, multi‐media models such as the US EPA's EpiSuite and ECOSAR are acceptable when measured data are not available. Uncertainty points are assigned based on the source of the data, whether they are measured or estimated or surrogate data. The final bioaccumulation chemical score is multiplied by the final persistence score and the result is then multiplied by a weighing factor of 1.4. The environmental fate of a chemical is emphasized in SCRAM because of the potential for a chemical deemed not toxic during laboratory studies later potentially to be found to cause toxicity through other mechanisms. A minimum of one data point in at least one toxicity category is required. Acute toxicity scores are composed of two components, acute aquatic and acute terrestrial toxicity, and are based on E/LC₅₀ values for aquatic organisms and E/LD₅₀ values for terrestrial organisms. Subchronic/chronic scores are based on LO(A)ELs and NO(A)ELs for aquatic and terrestrial organisms in addition to human toxicity values. The subchronic/chronic scores represent repeat dose toxicity, general organ toxicity and also carcinogenicity, mutagenicity, neurotoxicity, immunotoxicity, reproductive and developmental toxicity and endocrine disruption. The most conservative (or lowest) acute and subchronic/chronic data are used when scoring for toxicity. Both the acute toxicity score and the subchronic/chronic toxicity scores are added to uncertainty scores before being summed with the bioaccumulation and persistence scores for a final chemical score. The final chemical score is summed together with a final uncertainty score to give a final composite score that can be compared with those of other chemicals. The relative rankings of chemicals can aid scientists in determining which chemicals need more regulation and/or additional research.

The Chemicals Assessment and Ranking System (CARS) is a decision support tool developed by the Zero Waste Alliance (ZWA) that provides a process for inventorying, assessing and ranking chemicals.²²  Similarly to the SCRAM paradigm, chemicals are ranked according to their potential impacts on human health and safety, ecological health and ecosystem‐wide impacts. CARS has been used in support of environmental management systems such as those defined by ISO 14000. The first step is an inventory of chemicals used within an organization based on Material Safety Data Sheets (MSDS). The next step is to screen those chemicals against the CARS database to identify any suspected or potential carcinogens, teratogens, persistence bioaccumulative toxins, global warming gases, ozone‐depleting chemicals and more. The CARS database utilizes hazard lists from sources such as the US EPA, the American Conference of Governmental Industrial Hygienists (ACGIH), the National Toxicology Program (NTP) and the International Agency for Research on Cancer (IARC). A Prioritization Criteria Worksheet summarizes the results of the screen and provides a summary of chemicals of concern and the products that contain them. Products are then ranked based on their frequency of use, potential human health and safety impacts, ecological and global impacts and life‐cycle costs associated with storage, disposal, training and management. CARS is used to prioritize chemical products for replacement. Specific methods and criteria used in CARS to rank chemicals are not publicly accessible.

In 2001, the US company SC Johnson & Son (SCJ) developed a system known as Greenlist™ to classify ingredients found in SCJ products based on each chemical's impact on the environment and human health.²³  So far, SCJ has used Greenlist™ to rate over 95% of their products, and although the Greenlist™ criteria can be obtained from SCJ, the list of chemicals evaluated under Greenlist™ is not made public. The Greenlist™ process has been validated by the UK's Forum for the Future and the US EPA. Each potential ingredient receives a rating from 0 (Restricted Use) to 3 (Best). The company strives to continually improve its overall Greenlist™ ratings. According to SCJ, 27% of ingredients used in its products are classified as ‘best’ ingredients.²³  Greenlist™ chemical ratings are confidential, so individual chemical and product scores are not available. SCJ licenses the overall GreenList™ process to other companies free of charge,²³  but the total number of companies who have adopted GreenList™ and tailored this process to their own needs is unknown.

PRIO is an automated, web‐based tool intended to reduce the risks to human health and the environment from chemicals.²⁴  Developed by the Swedish Chemicals Agency, this model recommends phasing out high‐priority chemicals to achieve a non‐toxic environment. The tool is not intended to rate or score chemicals based on their human health and environmental hazards, rather it is used to identify the hazardous properties of a chemical. PRIO applies only to chemicals of high concern and categorizes them into two groups: phase‐out substances and priority risk‐reduction substances. Phase‐out substances are those that are of such high concern that they should not be used, such as PBT (Persistent, Bioaccumulative and Toxic) chemicals. Priority risk‐reduction substances are those to which special attention should be paid. The criteria used by PRIO have been based on REACH legislation and EU risk phrases. The only criteria against which phase‐out substances are evaluated are carcinogenicity, mutagenicity, reproductive toxicity and endocrine disruption. In addition to these, compounds are also assessed as to whether or not they contain any hazardous metals (mercury, cadmium and lead). Priority risk‐reduction substances are evaluated for acute and chronic toxicity, sensitization and mutagenicity. Both categories are evaluated for their environmental hazards: persistence and bioaccumulation. Phase‐out substances are also assessed for their ozone depletion properties.

The Quick Scan method was developed by the Dutch Ministry of Housing, Spatial Planning and the Environment to implement a chemicals substitution policy for chemicals with high hazards.²⁵  The Quick Scan attempts to develop chemical profiles based on hazard data, classify chemicals into categories of concern and assist industry in taking action for chemicals of high concern. The endpoints assessed during a Quick Scan include persistence (P), bioaccumulation (B), (eco)toxicity (T), health damage in humans (He), carcinogenicity (C), mutagenicity (M), reprotoxicity (R) and hormone disruptive effects (Ho). Hazard data on the chemical in question are gathered and then assessed in order to assign the chemical the appropriate hazard level. Chemicals are also categorized based on the level of concern: ‘Very High Concern,’ ‘High Concern,’ ‘Concern,’ ‘Low Concern,’ ‘No Data, Very High Concern.’ Concern categories are adjusted based on the likelihood of exposure. For example, the Quick Scan has four categories of exposure potential based on chemical use: intermediates, industrial applications, professional use and consumer use. Chemicals of Very High Concern are not to be used whereas substances of High Concern are not permitted for use in consumer products or in open profession use. Substances of Concern are permitted with limitations. The Quick Scan does not lead the user to a final selection, but only screens chemicals into broad categories based on a level of concern.

The Column Model was developed by the German Institute for Occupational Safety to comply with the German Hazardous Substances Ordinance that requires companies to replace hazardous substances with substances with a lower health risk. Since being introduced in 2001, the Column Model²⁶  has been updated to reflect GHS hazard classifications.²⁷  Similarly to the Dutch Quick Scan, chemical hazard data are presented in tabular form including six endpoints: acute health hazards, chronic health hazards, environmental hazards, fire and explosion hazards, exposure potential and hazards caused by procedures. Each chemical or ingredient is evaluated against each endpoint and assigned one of five hazard levels: very high, high, medium, low and negligible. Unlike the Quick Scan, the Column Method does not categorize chemicals based on levels of concern. The criteria used to assign a hazard level are determined primarily by risk phrases (R‐phrases) (for the original Column Model) or hazard phrases (H‐phrases) (for the newer GHS Column Model). Assessors are responsible for identifying and analyzing data. The Column Model and GHS Column Model are not comparative methods; however, the models can be used in two different ways: dominance analysis and positional analysis. In dominance analysis, two alternatives are compared and if one is ‘dominated’ by the other (i.e. poses a greater risk), the dominated alternative is discarded and another selected for comparison until one, non‐dominated alternative exceeds all those analyzed. In positional analysis, the decision is made based on the criteria considered most important to the user.

Developed for the German Federal Environmental Agency, the Evaluation Matrix is an aggregated data method, similar to the Column Method, which defines risk levels for specific endpoints and uses.²⁸  The Evaluation Matrix is a template that allows the assessor to perform an evaluation based on sustainability using substance‐specific criteria. Chemicals are evaluated using tables with specific indicators and the colors green, yellow, red and white are used to indicate the result. Eight substance‐specific criteria are evaluated and involve checking to see if that chemical is present on several lists of problematic substances, the possible dangerous physico‐chemical properties of the substance, human toxicity, problematic properties related to the environment, mobility within the environment, the origin of the raw materials used, emission of greenhouse gases and resource consumption. Information on the substance's mobility can be found in its MSDS such as EU risk phrases or in other publicly available resources. The criterion used when evaluating whether or not a substance is a PVT or vPvB chemical is that of REACH Annex XIII (see Tables 1.2 and 1.3). Following the evaluation against these endpoints, substances are placed into one of four categories: Green (No action is needed, because available information indicates that the chemical is not problematic), Yellow (No action is needed, because the available information indicates problematic substance properties), Red (There is a high priority to act, because the available information indicates very problematic substance properties), White (There is a need to gather further information, because no or few data are available). A risk index can be created by weighting the endpoints and then summing up the weighted values.

Several drawbacks exist when conducting a CAA. The primary challenge is associated with the correct management of hazard and other trade‐offs. Switching from a chemical that poses a moderate human health risk to an alternative whose degradation products are aquatically toxic would not be a desirable substitution. This is one example how using CHA and life‐cycle thinking would be beneficial when conducting a CAA. It is important to note that although several of the above‐mentioned CAA methods use criteria that are more in‐depth, not all CAA methods take into account parameters such as resource consumption, energy usage or recyclability. Another primary challenge is complete characterization of the hazard profile for a chemical that has an incomplete data set for human health effects or environmental fate and toxicity.

In instances where data gaps exist for a chemical (either for a health effects or environmental effects endpoint), it is sometimes possible to characterize the potential hazard of that chemical by evaluating data on chemical surrogates or using software to predict the chemical's potential hazard.

Hazard characterization data gaps can often be addressed by evaluating hazard data pertaining to one or more structurally similar surrogates. This approach is based on the assumption that a chemical's structure imparts properties that relate to biological activity and that a group of chemicals that produce the same activity have something similar about their chemistry and/or structure. According to the OECD guidelines, an analog selected to fill a data gap must be data rich and share similar physical and chemical properties, including behavior in physical or biological process, with the original compound.²⁹  Chemicals produced by similar methods by the same company and used for similar purposes make good potential analogs. In addition, degradation products of the parent compound can be used as surrogates, especially if the parent compound is expected to break down readily in the environment.

The US EPA (2010)³⁰  and OECD (2007)²⁹  have defined guidelines for identifying similar substances to use analogs based on the following commonalities:

• A common functional group or substance (e.g. phenols, aldehydes).

• A common precursor or breakdown product may result in structurally similar chemicals, which can be used to examine related chemicals such as acids/esters/salts (e.g. short‐chain alkyl methacrylate esters which are metabolized to methacrylic acid).

• An incremental or constant change (e.g. increased carbon chain length; typically used for physico‐chemical properties such as boiling point).

• Common constituents or chemical class, similar carbon range numbers – used with substances of unknown or variable composition, complex reaction products or biological material.

At all times, the practitioner must include the rationale for his or her choice of analog(s) in the CAA.

If a structurally similar analog is not available, a modeling software program may be suitable to satisfy any data gaps. These computerized systems predict toxicity using structure–activity relationships. Examples of software programs used in CAAs include (but are not limited to) the following:

• OncoLogic (carcinogenicity): http://www.epa.gov/oppt/sf/pubs/oncologic.htm

• Toxicity Estimation Software Tool (TEST): http://www.epa.gov/nrmrl/std/qsar/qsar.html#TEST

• Estimation Program Interface (EPI) Suite (environmental fate): http://www.epa.gov/oppt/exposure/pubs/episuite.htm

• Ecological Structure–Activity Relationships (ECOSAR) (aquatic toxicity): http://www.epa.gov/oppt/newchems/tools/21ecosar.htm

• ToxTree (toxic hazard estimation): http://toxtree.sourceforge.net/

• OECD QSAR Toolbox (www.qsartoolbox.org)

• VEGA (Virtual Models for Evaluating the Properties of Chemicals within a Global Architecture) (www.vega‐qsar.eu/download.html).

All output/estimates generated from modeling programs should be appended to a CAA to promote transparency, accuracy and accountability. In some cases, data gaps cannot be filled because viable analogs are not available and the models may not be appropriate. In such situations, data gaps must be clearly presented and weighted in the assessment.

Finding safer alternatives to problematic chemicals is a growing global concern. The development of frameworks, tools and paradigms has been fueled by regulatory and non‐regulatory initiatives at all levels, such as REACH in the European Union and the widespread adoption of the United Nation's GHS around the world. Ideally, a CAA supports the intelligent creation, use and substitution of chemicals to benefit humankind in manners that will not harm the environment or organisms inhabiting the environment.

A number of the CAA methods described in this chapter have advantages over others. Some CAA paradigms place an emphasis on human health hazards, whereas others only address environmental hazards. None of the CAA methods in this chapter are yet automated and require 30–60h of highly technical work per chemical to characterize its potential hazard. CAA methods require the evaluator to be skilled in toxicology, chemistry, ecotoxicology and environmental science, in addition to having a working knowledge of LCA methods and concepts. Not all CAA paradigms consider LCA attributes. As the complexities of the substance being evaluated (chemical, material, product) increase, so do the complexities of the CAA evaluation. The best CAA paradigms are those that are flexible, adhere strictly to transparency and can be modified in order to meet the specific goals of the evaluator.

Improvement in CAA requires greater transparency in the actual methods employed as part of a CAA. Of the CCA methods discussed in this chapter, only the US EPA's DfE Alternatives Assessment, CPA's GreenScreen™ and Cradle to Cradle® have publicly available criteria and background materials. Most of the CAA methodologies were developed a decade ago, so it is likely that hundreds if not thousands of CAAs have been performed using one of the CAA methods described in this chapter. Despite this, there is no central database that can be accessed to search for completed CAAs. An online database that could index completed CAAs using different CAA paradigms would save time and resources by minimizing duplication of effort among CAA assessors who currently end up performing CAAs that other organizations have likely assessed under one or more of the prevailing CAA methods.

To date, a chemical's human health or environmental footprint has taken a backseat to attributes such as a chemical's impact on a product's performance, reliability or price in the marketplace. This way of thinking is not sustainable or preferable, as only a portion of a chemical's true cost and impact are considered—or paid for—in the marketplace. Similarly, banning or restricting chemicals on an ad hoc basis is not a solution to our current system, nor is a system of positive lists allowing the use of only certain chemicals, as that stifles innovation and continuous improvement. Early measures such as the Montreal Protocol and the Basel Convention, and more recently REACH and GHS, demonstrate that faults in the current system are recognized; however, these initiatives do not instill a fundamental change in our way of thinking. CAAs provide a powerful means to improve upon the status quo by establishing methods to inform chemical substitution in a scientifically rigorous and defensible manner. Obviously, there is great room for refinement and improvement in CAA methods, as this is a relatively young discipline. It is not the nature of humans−or any living entity−to start out by giving up.¹⁵  Recognizing the value of CAA and fostering greater adoption of CAA methods provide stakeholders with much‐needed tools to address a serious deficiency in the way in which chemicals are used in society, as maintaining the status quo is analogous to admitting defeat. As humankind's understanding of the full costs and benefits of chemicals matures, it is critical that we cease using those chemicals that can permanently impair human health or the environment.