United States Environmental Protection Agency EPA Document# EPA-740-R1-7007 June 2017 Office of Chemical Safety and Pollution Prevention Scope of the Risk Evaluation for Perchloroethylene (Ethene, 1,1,2,2-Tetrachloro) CASRN: 127-18-4 June 2017 are provided in Table 2-2 from EPA’s CDR database (U.S. EPA, 2016b). As demonstrated in Table 2-2, production and importation of perchloroethylene has decreased since 2012. Table 2-2. Production Volume of Perchloroethylene in CDR Reporting Period (2012 to 2015) a Reporting Year Total Aggregate Production Volume (lbs) 2012 2013 2014 2015 387,623,401 391,403,540 355,305,850 324,240,744 a The CDR data for the 2016 reporting period is available via ChemView (https://java.epa.gov/chemview) (U.S. EPA, 2016b). Because of an ongoing CBI substantiation process required by amended TSCA, the CDR data available in the scope document is more specific than currently in ChemView. Figure 2-1 depicts the initial life cycle diagram of perchloroethylene from manufacture to the point of disposal. EPA identified the use categories by reviewing the industrial processing use activities, and commercial and consumer use product categories reported in the 2016 CDR (U.S. EPA, 2016b). Then, EPA identified the subcategories by supplementing CDR data with information from Preliminary Information on Manufacturing, Processing, Distribution, Use, and Disposal: Tetrachloroethylene (Perchloroethylene) and Use and Market Profile for Tetrachloroethylene, both available in the public docket (EPA-HQ-OPPT-2016-0732). For risk evaluations, EPA will assess each use subcategory by identifying all potential sources of release and human exposure associated with that subcategory. Page 20 of 77 INTERNAL DELIBERATIVE - DO NOT QUOTE OR CITE Figure 2-1. Initial Perchloroethylene Life Cycle Diagram The initial life cycle diagram depicts the conditions of use that are within the scope of the risk evaluation during various life cycle stages including manufacturing, processing, use (industrial, commercial, consumer, where distinguishable), distribution and disposal. The production volumes shown are for reporting year 2015 from the 2016 CDR reporting period (U.S. EPA, 2016b). Activities related to distribution (e.g., loading, unloading) will be considered throughout the perchloroethylene life cycle, rather than using a single distribution scenario. a See Table 2-3 for additional uses not mentioned specifically in this diagram. Wastewater: combination of water and organic liquid, where the organic content is less than 50 %. Liquid Wastes: combination of water and organic liquid, where the organic content is greater than 50 %. b Page 21 of 77 DRAFTUnited States Environmental Protection Agency EPA Document# EPA-740-R1-7017 May 2018 Office of Chemical Safety and Pollution Prevention Problem Formulation of the Risk Evaluation for Perchloroethylene (Ethene, 1,1,2,2-Tetrachloro) CASRN: 127-18-4 May 2018 EXECUTIVE SUMMARY TSCA § 6(b)(4) requires the U.S. Environmental Protection Agency (EPA) to establish a risk evaluation process. In performing risk evaluations for existing chemLFDOV (3$ LV GLUHFWHG WR ³GHWHUPLQH ZKHWKHU D chemical substance presents an unreasonable risk of injury to health or the environment, without consideration of costs or other non-risk factors, including an unreasonable risk to a potentially exposed or susceptible subpopulation identified as relevant to the risk evaluation by the Administrator under the FRQGLWLRQV RI XVH ´ ,Q 'HFHPEHU RI (3$ SXEOLVKHG D OLVW RI chemical substances that are the VXEMHFW RI WKH $JHQF\¶V LQLWLDO FKHPLFDO ULVN HYDOXDWLRns (81 FR 91927), as required by TSCA § 6(b)(2)(A). Perchloroethylene was one of these chemicals. TSCA § 6(b)(4)(D) requires that EPA publish the scope of the risk evaluation to be conducted, including the hazards, exposures, conditions of use and potentially exposed or susceptible subpopulations that the Administrator expects to consider. In June 2017, EPA published the Scope of the Risk Evaluation for perchloroethylene. As explained in the scope document, because there was insufficient time for EPA to provide an opportunity for comment on a draft of the scope, as EPA intends to do for future scope documents, EPA is publishing and taking public comment on a problem formulation document to refine the current scope, as an additional interim step prior to publication of the draft risk evaluation for perchloroethylene. Comments received on this problem formulation document will inform development of the draft risk evaluation. This problem formulation document refines the conditions of use, exposures and hazards presented in the scope of the risk evaluation for perchloroethylene and presents refined conceptual models and analysis plans that describe how EPA expects to evaluate the risk for perchloroethylene. Perchloroethylene, also known as ethene, 1,1,2,2-tetrachloro, tetrachloroethylene and PCE, is a high production volume (HPV) solvent. Perchloroethylene is subject to a number of federal and state regulations and reporting requirements. For example, perchloroethylene has been a Toxics Release Inventory (TRI) reportable chemical under Section 313 of the Emergency Planning and Community Right-to-Know Act (EPCRA) since 1995. It is designated a Hazardous Air Pollutant (HAP) under the Clean Air Act (CAA), a hazardous waste under the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA) and a regulated drinking water contaminant under the Safe Drinking Water Act (SDWA). Information on the domestic manufacture, processing and use of perchloroethylene is available to EPA through its Chemical Data Reporting (CDR) Rule, issued under TSCA. According to the 2016 CDR, more than 324 million pounds of perchloroethylene were manufactured (including imported) in the United States in 2015. According to the Use and Market Profile for Tetrachloroethylene (EPA-HQOPPT-2016-0732), perchloroethylene is primarily used to produce fluorinated compounds, such as hydrofluorocarbons (HFCs) and hydrochlorofluorocarbons (HCFCs) (65%) followed by dry cleaning (15%) and vapor degreasing solvents (10%). Other uses can be quite varied, including: x Adhesives x Degreasing x Brake cleaner x Laboratories x Lubricants x Mold cleaners, releases and protectants x Oil refining WĂŐĞ ϭϭ ŽĨ ϭϲϳ x x x x Sealants Stainless steel polish Tire buffers and cleaners and Vandal mark removers. This document presents the potential exposures that may result from the conditions of use of perchloroethylene. Exposures may occur to workers and occupational non-users (workers who do not directly handle the chemical but perform work in an area where the chemical is used), consumers and bystanders (non-product users that are incidentally exposed to the product) and the general population through inhalation, dermal and oral pathways. Workers and occupational non-users (ONU), who do not directly handle the chemical but perform work in an area where the chemical is used, may be exposed to perchloroethylene during a variety of conditions of use, such as manufacturing, processing and industrial and commercial uses, including uses in degreasing and adhesives. EPA expects that the highest exposures to perchloroethylene generally involve workers in industrial and commercial settings. Perchloroethylene can be found in numerous products and can, therefore, result in exposures to commercial and consumer users in indoor or outdoor environments. For perchloroethylene, EPA considers workers, occupational non-users, consumers, bystanders, and certain other groups of individuals who may experience greater exposures than the general population due to proximity to conditions of use to be potentially exposed or susceptible subpopulations. Exposures to the general population may occur from industrial and/or commercial uses; industrial releases to air, water or land; and other conditions of use. EPA will evaluate whether groups of individuals within the general population may be exposed via pathways that are distinct from the general population due to unique characteristics (e.g., life stage, behaviors, activities, duration) that increase exposure and whether groups of individuals have heightened susceptibility, and should therefore be considered potentially exposed or susceptible subpopulations for purposes of the risk evaluation. EPA plans to further analyze inhalation exposures to vapors and mists for workers and occupational non-users and dermal exposures for skin contact with liquids in occluded situations for workers in the risk evaluation. For environmental release pathways, EPA plans to further analyze surface water exposure to aquatic vertebrates, invertebrates and aquatic plants and exposure to sediment-dwelling organisms. Perchloroethylene has been the subject of several prior health hazard and risk assessments, including (3$¶V Integrated Risk Information System (IRIS) Toxicological Review and a draft Agency for Toxic Substances and DiseaVH 5HJLVWU\¶V $76'5¶V 7R[LFRORJLFDO 3URILOH A number of targets of toxicity from exposures to perchloroethylene have been identified in animal and human studies for both oral and inhalation exposures. EPA plans to evaluate all potential hazards for perchloroethylene, using the primary literature identified in human health reviews and including any found in recent literature. Hazard endpoints identified in previous assessments include: acute toxicity, neurotoxicity, kidney toxicity, liver toxicity, developmental and reproductive toxicity and cancer. Support for an association with immune and blood effects was less well characterized. Perchloroethylene is also considered to be irritating. The revised conceptual models presented in this problem formulation identify conditions of use; exposure pathways (e.g., media); exposure routes (e.g., inhalation, dermal, oral); potentially exposed or susceptible subpopulations; and hazards EPA expects to consider in the risk evaluation. The initial conceptual models provided in the scope document were revised during problem formulation based on evaluation of reasonably available information for physical and chemical properties, fate, exposures, hazards and conditions of use, and based upon consideration of other statutory and regulatory authorities. In each problem formulation document for the first 10 chemical substances, EPA also WĂŐĞ ϭϮ ŽĨ ϭϲϳ refined the activities, hazards and exposure pathways that will be included in and excluded from the risk evaluation. (3$¶V RYHUDOO REMHFWLYHV LQ WKH ULVN HYDOXDWLRQ SURFHVV DUH WR FRQGXFW WLPHO\ UHOHYDQW KLJK-quality, and scientifically credible risk evaluations within the statutory deadlines, and to evaluate the conditions of use that raise greatest potential for risk 82 FR 33726, 33728 (July 20, 2017). WĂŐĞ ϭϯ ŽĨ ϭϲϳ 1 INTRODUCTION This document presents for comment the problem formulation of the risk evaluation to be conducted for perchloroethylene under the Frank R. Lautenberg Chemical Safety for the 21st Century Act. The Frank R. Lautenberg Chemical Safety for the 21st Century Act amended the Toxic Substances Control Act 76&$ WKH QDWLRQ¶V SULPDU\ FKHPLFDOV PDQDJHPHQW ODZ RQ -XQH 7KH QHZ ODZ LQFOXGHV statutory requirements and deadlines for actions related to conducting risk evaluations of existing chemicals. In December of 2016, EPA published a list of 10 chemical substances that are the subject of the $JHQF\¶V LQLWLDO FKHPLFDO ULVN HYDOXDWLRQV 81 FR 91927), as required by TSCA § 6(b)(2)(A). These 10 chemical substances werH GUDZQ IURP WKH XSGDWH RI (3$¶V 76&$ :RUN 3ODQ IRU &KHPLFDO Assessments, a list of chemicals that EPA identified in 2012 and updated in 2014 (currently totaling 90 FKHPLFDOV IRU IXUWKHU DVVHVVPHQW XQGHU 76&$ (3$¶V GHVLJQDWLRQ RI WKH ILUVW chemical substances constituted the initiation of the risk evaluation process for each of these chemical substances, pursuant to the requirements of TSCA § 6(b)(4). TSCA § 6(b)(4)(D) requires that EPA publish the scope of the risk evaluation to be conducted, including the hazards, exposures, conditions of use and potentially exposed or susceptible subpopulations that the Administrator expects to consider, within 6 months after the initiation of a risk evaluation. The scope documents for all first 10 chemical substances were issued on June 22, 2017. The first 10 problem formulation documents are a refinement of what was presented in the first 10 scope documents. TSCA § 6(b)(4)(D) does not distinguish between scoping and problem formulation, and requires EPA to issue scope documents that include information about the chemical substance, including the hazards, exposures, conditions of use, and the potentially exposed or susceptible subpopulations that the Administrator expects to consider in the risk evaluation. In the future, EPA expects scoping and problem formulation to be completed prior to the issuance of scope documents and intends to issue scope documents that include problem formulation. As explained in the scope document, because there was insufficient time for EPA to provide an opportunity for comment on a draft of the scope, as EPA intends to do for future scope documents, EPA is publishing and taking public comment on a problem formulation document to refine the current scope, as an additional interim step prior to publication of the draft risk evaluation for perchloroethylene. Comments received on this problem formulation document will inform development of the draft risk evaluation. The Agency defines problem formulation as the analytical phase of the rLVN DVVHVVPHQW LQ ZKLFK ³WKH purpose for the assessment is articulated, the problem is defined, and a plan for analyzing and FKDUDFWHUL]LQJ ULVN LV GHWHUPLQHG´ VHH Section 2.2 of the Framework for Human Health Risk Assessment to Inform Decision Making). The outcome of problem formulation is a conceptual model(s) and an analysis plan. The conceptual model describes the linkages between stressors and adverse human health effects, including the stressor(s), exposure pathway(s), exposed life stage(s) and population(s), and endpoint(s) that will be addressed in the risk evaluation (U.S. EPA, 2014e). The analysis plan follows the development of the conceptual model(s) and is intended to describe the approach for conducting the risk evaluation, including its design, methods and key inputs and intended outputs as described in the EPA Human Health Risk Assessment Framework (U.S. EPA, 2014e). The problem formulation documents refine the initial conceptual models and analysis plans that were provided in the scope documents. WĂŐĞ ϭϰ ŽĨ ϭϲϳ First, EPA has removed from the risk evaluation any activities and exposure pathways that EPA has concluded do not warrant inclusion in the risk evaluation. For example, for some activities which were listed as "conditions of use" in the scope document, EPA has insufficient information following the further investigations during problem formulation to find they are circumstances under which the chemical is actually "intended, known, or reasonably foreseen to be manufactured, processed, distributed in commerce, used, or disposed of." Second, EPA also identified certain exposure pathways that are under the jurisdiction of regulatory programs and associated analytical processes carried out under other EPA-administered environmental statutes ± namely, the Clean Air Act (CAA), the Safe Drinking Water Act (SDWA), the Clean Water Act (CWA), and the Resource Conservation and Recovery Act (RCRA) ± and which EPA does not expect to include in the risk evaluation. As a general matter, EPA believes that certain programs under other Federal environmental laws adequately assess and effectively manage the risks for the covered exposure pathways. To use Agency resources efficiently under the TSCA program, to avoid duplicating efforts taken pursuant to other Agency programs, to maximize scientific and analytical efforts, and to meet the three-year statutory deadline, EPA is planning to exercise its discretion under TSCA 6(b)(4)(D) to focus its analytical efforts on exposures that are likely to present the greatest concern and consequently merit a risk evaluation under TSCA, by excluding, on a case-by-case basis, certain exposure pathways that fall under the jurisdiction of other EPA-administered statutes.1 EPA does not expect to include tany such excluded pathways as further explained below in the risk evaluation. The provisions of various EPA-administered environmental statutes and their implementing regulations represent the judgment of Congress and the Administrator, respectively, as to the degree of health and environmental risk reduction that is sufficient under the various environmental statutes. Third, EPA identified any conditions of use, hazards, or exposure pathways which were included in the scope document and that EPA expects to include in the risk evaluation but which EPA does not expect to further analyze in the risk evaluation. EPA expects to be able to reach conclusions about particular conditions of use, hazards or exposure pathways without further analysis and therefore expects to conduct no further analysis on those conditions of use, hazards or exposure pathways in order to focus WKH $JHQF\¶V UHVRXUFHV RQ PRUH H[WHQVLYH RU TXDQWLWDWLYH DQDO\ses. Each risk evaluation will be "fit-forpurpose," meaning not all conditions of use will warrant the same level of evaluation and the Agency may be able to reach some conclusions without comprehensive or quantitative risk evaluations 82 FR 33726, 33734, 33739 (July 20, 2017). EPA received comments on the published scope document for perchloroethylene and has considered the comments specific to perchloroethylene in this problem formulation document. EPA is soliciting public comment on this problem formulation document and when the draft risk evaluation is issued the Agency intends to respond to comments that are submitted. In its draft risk evaluation, EPA may revise the conclusions and approaches contained in this problem formulations, including the conditions of use and pathways covered and the conceptual models and analysis plans, based on comments received. $V H[SODLQHG LQ WKH ILQDO UXOH IRU FKHPLFDO ULVN HYDOXDWLRQ SURFHGXUHV ³EPA may, on a case-by case basis, exclude certain activities that EPA has determined to be conditions of use in order to focus its analytical efforts on those exposures that are likely to present the greatest concern, and consequently merit an unreasonable risk determination ´ [82 FR 33726, 33734, 33729 (July 20, 2017)] 1 WĂŐĞ ϭϱ ŽĨ ϭϲϳ 2.2 Conditions of Use 76&$ † GHILQHV WKH FRQGLWLRQV RI XVH DV µµWKH FLUFXPVWDQFHV DV determined by the Administrator, under which a chemical substance is intended, known, or reasonably foreseen to be manufactured, SURFHVVHG GLVWULEXWHG LQ FRPPHUFH XVHG RU GLVSRVHG RI ¶¶ Data and Information Sources In the scope documents, EPA identified, based on reasonably available information, the conditions of use for the subject chemicals. As further described in this document, EPA searched a number of available data sources (e.g., Use and Market Profile for Tetrachloroethylene, EPA-HQ-OPPT-20160732). Based on this search, EPA published a preliminary list of information and sources related to chemical conditions of use [see Preliminary Information on Manufacturing, Processing, Distribution, Use, and Disposal: Tetrachloroethylene (Perchloroethylene) and Use, EPA-HQ-OPPT-2016-0732] prior to a February 2017 public meeting on scoping efforts for risk evaluation convened to solicit comment and input from the public. EPA also convened meetings with companies, industry groups, chemical users and other stakeholders to aid in identifying conditions of use and verifying conditions of use identified by EPA. The information and input received from the public and stakeholder meetings has been incorporated into this problem formulation document to the extent appropriate. Thus, EPA believes the manufacture, processing, distribution, use and disposal activities identified in these documents constitute the intended, known, and reasonably foreseeable activities associated with the subject chemical, based on reasonably available information. Identification of Conditions of Use To determine the current conditions of use of perchloroethylene and inversely, activities that do not qualify as conditions of use (3$ FRQGXFWHG H[WHQVLYH UHVHDUFK DQG RXWUHDFK 7KLV LQFOXGHG (3$¶V review of published literature and online databases including the most recent data avaLODEOH IURP (3$¶V Chemical Data Reporting program (CDR) and Safety Data Sheets (SDSs). EPA also conducted online research by reviewing company websites of potential manufacturers, importers, distributors, retailers, or other users of perchloroethylene and queried government and commercial trade databases. EPA also received comments on the Scope of the Risk Evaluation for perchloroethylene (EPA-HQ-OPPT-20160732) that were used to determine the conditions of use. In addition, EPA convened meetings with companies, industry groups, chemical users, states, environmental groups, and other stakeholders to aid in identifying conditions of use and verifying conditions of use identified by EPA. Those meetings included a February 14, 2017 public meeting with such entities (EPA-HQ-OPPT-2016-0732). EPA has removed from the risk evaluation anyy activities that EPA concluded do not constitute conditions of use ± for example p because EPA has insufficient information to find certain activities are FLUFXPVWDQFHV XQGHU ZKLFK WKH FKHPLFDO LV DFWXDOO\ ³LQWHQGHG NQRZQ RU UHDVRQDEO\ IRUHVHHQ WR EH \ N \ manufaFWXUHG SURFHVVHG GLVWULEXWHG LQ FRPPHUFH XVHG RU GLVSRVHG RI ´ (3$ KDV DOVR LGHQWLILHG DQ\ S S I \ conditions of use that EPA does not expect p to include in the risk evaluation. As explained p in the final rule for Procedures for Chemical Risk Evaluation Under the Amended Toxic Substances Control Act,, TSCA Section 6(b)(4)(D) ( )( )( ) requires q EPA to identifyy "the hazards,, exposures, p , conditions of use,, and the or susceptible subpopulations ppotentially y exposed p p p p the Administrator expects p to consider´ in a risk evaluation,, suggesting gg g that EPA is not required q to consider all conditions of use,, and EPA may exclude certain activities that EPA has determined to be conditions of use on a case-by-case basis 82 FR 33736, 33729 (July 20, 2017). For example, EPA may exclude conditions of use that the Agency has sufficient basis to conclude would present only de minimus exposures or otherwise insignificant risks (such as use in a closed system that effectively precludes exposure or as an intermediate). WĂŐĞ Ϯϭ ŽĨ ϭϲϳ The activities that EPA no longer believes are conditions of use or were otherwise excluded during problem formulation are described in Section 2.2.2.1. The conditions of use included in the scope of the risk evaluation are summarized in Section 2.2.2.2. 2.2.2.1 Categories and Subcategories Determined Not to be Conditions of Use During Problem Formulation For perchloroethylene, EPA has conducted public outreach and literature searches to collect information about perchloroethylene's conditions of use and has reviewed reasonably available information obtained or possessed by EPA concerning activities associated with perchloroethylene. Based on the foregoing research and outreach, EPA does not have reason to believe that any categories or subcategories identified in the perchloroethylene scope should be excluded from the scope of the risk evaluation. Therefore, no categories or subcategories of use for perchloroethylene will be excluded from the scope of the risk evaluation. Table 2-2. Categories and Subcategories Determined Not to be Conditions of Use During Problem Formulation Life Cycle Stage Category a Subcategory b References No categories or subcategories have been excluded from the risk evaluation. 2.2.2.2 Categories and Subcategories of Conditions of Use Included in the Scope of the Risk Evaluation The uses of perchloroethylene include the production of fluorinated compounds, dry cleaning and vapor degreasing, as well as a number of smaller uses. Nearly 65% of the production volume of perchloroethylene is used as an intermediate in industrial gas manufacturing, more specifically to produce fluorinated compounds, such as hydrofluorocarbons (HFCs) and hydrochlorofluorocarbons (HCFCs) (NTP, 2014; ICIS, 2011). HFCs 134a and 125 are alternatives to chlorofluorocarbons (CFCs) and HCFCs, which are ozone depleting substances (ODSs), and the subject of a phase-out (https://www.epa.gov/ods-phaseout). HCFCs are transitional substances in the phase-out of ODSs (ICIS, 2011) (Public Comment, EPA-HQ-OPPT-2016-0732-0033). Previously, perchloroethylene was widely used to manufacture CFCs (esp. trichlorotrifluoroethane (CFC-113)) until production and importation of CFCs for most uses were phased out in the United States by regulations implementing the Montreal Protocol (40 CFR part 82). A relatively small amount of CFC-113 is still produced for exempted uses (teleconference with Honeywell, 2017; summary is available in the docket: EPA-HQ-OPPT-2016-0732). The second largest use of perchloroethylene (~15%) is as a solvent in dry cleaning facilities (NTP, 2014). Perchloroethylene is non-flammable and effectively dissolves fats, greases, waxes and oils, without harming natural or human-made fibers. These properties enabled it to replace traditional petroleum solvents(ATSDR, 2014; Dow Chemical Co, 2008; Tirsell, 2000). The demand for perchloroethylene dry cleaning solvents has steadily declined as a result of the improved efficiency of dry cleaning equipment, increased chemical recycling and the popularity of wash-and-wear fabrics that eliminate the need for dry cleaning (ATSDR, 2014). Perchloroethylene is also used in dry cleaning detergent and dry cleaning sizing. WĂŐĞ ϮϮ ŽĨ ϭϲϳ Approximately 60% of dry cleaning machines now use perchloroethylene as a solvent (DLI and NCA, 2017). In 1991, EPA estimated that 83% of all dry cleaning facilities used perchloroethylene as solvent (U.S. EPA, 1991). In 2008, the Halogenated Solvents Industry Association (HSIA) estimated that 70% of dry cleaners used perchloroethylene as dry cleaning solvent (EPA-HQ-OPPT-2016-0732-0027). Similarly, in 2011, King County, WA conducted a profile of the dry cleaning industry and found that 69% of respondents (105 of the 152 respondents) used perchloroethylene in their primary machine (Whittaker and Johanson, 2011). Hence, there appears to be a trend towards alternatives to perchloroethylene in dry cleaning. According to the dry cleaning industry, a majority of new perchloroethylene dry cleaning machines are sold in locations where local fire codes preclude the use of Class III combustible alternative solvents or where the nature of the dry cleaning operation requires the use of perchloroethylene (DLI and NCA, 2017). The third most prevalent use of perchloroethylene (~10%) is as a vapor degreasing solvent (NTP, 2014). Perchloroethylene can be used to dissolve many organic compounds, select inorganic compounds and high-melting pitches and waxes making it ideal for cleaning contaminated metal parts and other fabricated materials (ATSDR, 2014). It is a very good solvent for greases, fats, waxes, oils, bitumen, tar and many natural and synthetic resins for use in chemical cleaning systems, degreasing light and heavy metals, degreasing pelts and leather (tanning), extraction of animal and vegetable fats and oils and textile dyeing (solvent for dye baths)(Stoye, 2000). Perchloroethylene is also used in cold cleaning, which is similar to vapor degreasing, except that cold cleaning does not require the solvent to be heated to its boiling point in order to clean a given component. Vapor degreasing and cold cleaning scenarios may include a range of open-top or closed systems, conveyorized/enclosed/inline systems, spray wands, dip containers and wipes. Perchloroethylene has many other uses, which collectively constitute ~10% of the production volume. (3$¶V VHDUFK RI VDIHW\ GDWD VKHHWV JRYHUQPHQW GDWDEDVHV DQG RWKHU VRXUFHV IRXQG RYHU products containing perchloroethylene. These uses include (but are not limited to): x Adhesives x Aerosol degreasing x Brake cleaner x Laboratories x Lubricants x Mold cleaners, releases and protectants x Oil refining x Sealants x Stainless steel polish x Tire buffers and cleaners x Vandal mark removers Many of these uses include consumer products, such as adhesives (arts and crafts, as well as light repairs), aerosol degreasing, brake cleaners, aerosol lubricants, sealants, sealants for gun ammunition, stone polish, stainless steel polish and wipe cleaners. The uses of perchloroethylene in consumer adhesives and brake cleaners are especially prevalent; EPA has found 16 consumer adhesive products and 14 consumer brake cleaners containing perchloroethylene [see Preliminary Information on Manufacturing, Processing, Distribution, Use, and Disposal: Tetrachloroethylene (Perchloroethylene) and Use and Market Profile for Tetrachloroethylene, EPA-HQ-OPPT-2016-0732-0003]. WĂŐĞ Ϯϯ ŽĨ ϭϲϳ Table 2-3 summarizes each life cycle stage and the corresponding categories and subcategories of conditions of use for perchloroethylene that EPA expects to consider in the risk evaluation. Using the 2016 CDR (U.S. EPA, 2016b), EPA identified industrial processing or use activities, industrial function categories and commercial and consumer use product categories. EPA identified the subcategories by supplementing CDR data with other published literature and information obtained through stakeholder consultations. For risk evaluations, EPA intends to consider each life cycle stage (and corresponding use categories and subcategories) and assess certain relevant potential sources of release and human exposure associated with that life cycle stage. Beyond the uses identified in the Scope of the Risk Evaluation for Perchloroethylene, EPA has received no additional information identifying additional current conditions of use for perchloroethylene from public comment and stakeholder meetings. WĂŐĞ Ϯϰ ŽĨ ϭϲϳ Table 2-9: Ecological Hazard Characterization of Perchloroethylene Test organism Aquatic Organisms Fish Duration Acute Endpoint Hazard value* Units Effect Endpoint LC50 4 ± 28.1 mg/L Mortality Aquatic invertebrates LC/EC50 2.85 ± 30.8 mg/L Algae Amphibians Acute COC EC50 3.64 - 500 EC50 2.5 -20.0 0.80 mg/L mg/L mg/L ChV 0.5-1.4 mg/L Mortality/ Reproduction Chronic Smith (1991); Horne (1983) Immobilization Hollister (1968); Call (1983) as cited in WHO (2006) Biomass/ Brack (1994) as cited in Abundance ECB (2005); U.S. EPA (1980a) as cited in WHO (2006) Mortality McDaniel (2004) Growth Fish References Aquatic 0.37 ± 1.11 invertebrates ChV (NOEC) mg/L NOEC 0.01-0.02 Abundance Algae LOEC 0.02-0.05 0.014-0.032 mg/L ChV Chronic COC 0.001 mg/L Terrestrial Organisms Terrestrial Cocoons invertebrates LC50 100 - 320 mg/Kg appearance Acute Terrestrial Growth EC50 861 mg/Kg plants Terrestrial Biomass Chronic plants EC50 12 mg/L * Values in the tables are presented as reported by the study authors Ahmad (1984); Smith (1991) as cited in ECB (2005) Hollister (1968); Richter et al. (1983) as cited in ECB (2005); Call (1983) as cited in WHO (2006) Labra (2010); (Vonk et al., 1986) as cited in WHO (2006) (Bauer and Dietze, 1992) as cited in WHO (2006) Hulzebos, 1993 Concentrations of Concern The screening-level acute and chronic concentrations of concern (COCs) for perchloroethylene were derived based on the lowest or most toxic ecological toxicity values (e.g., L/EC50). The information EHORZ GHVFULEHV KRZ WKH DFXWH DQG FKURQLF &2&¶V ZHUH FDOFXODWHG IRU HQYLURQPHQWDO WR[LFLW\ RI perchloroethylene using assessment factors. The application of assessment factors is based on established EPA/OPPT methods (U.S. EPA, 2013, 2012c) and were used in this hazard assessment to calculate lower bound effect levels (referred to as the concentration of concern; COC) that would likely encompass more sensitive species not specifically WĂŐĞ ϱϬ ŽĨ ϭϲϳ represented by the available experimental data. Also, assessment factors are included in the COC calculation to account for differences in inter- and intra-species variability, as well as laboratory-to-field variability. It should be noted that these assessment factors are dependent upon the availability of datasets that can be used to characterize relative sensitivities across multiple species within a given taxa or species group, but are often standardized in risk assessments conducted under TSCA, due to limited data availability. The concentrations of concern for each endpoint were derived based on the ecological hazard data for perchloroethylene. The information below describes how the acute and chronic COCs were calculated for aquatic toxicity. The acute COC is derived by dividing acute aquatic invertebrates LC50 of 2.85 mg/L (the lowest acute value in the dataset) by an assessment factor (AF) of 5: ‡ Lowest value for aquatic invertebrates LC50 (2.85 mg/L) / AF of 5 = 0.57 mg/L or 570 —g/L. The acute COC of 570 —g/L, derived from experimental aquatic LQYHUWHEUDWH¶V HQGSRLQW LV XVHG DV D conservative hazard level in this problem formulation for perchloroethylene. The chronic COC was determined based on the lowest chronic toxicity value divided by an assessment factor of 10. ‡ Lowest chronic value for 72-h algal ChV = 0.014 mg/L / 10 = 0.0014 mg/L or 1.4 —g/L. The chronic COC of 1.4 —g/L, derived from experimental algae endpoint, is used as the lower bound hazard level in this problem formulation for perchloroethylene. Human Health Hazards Perchloroethylene has an existing EPA IRIS Assessment U.S. EPA (2012e) and a draft ATSDR Toxicological Profile (ATSDR, 2014); hence, many of the hazards of perchloroethylene have been previously compiled. EPA expects to use these previous analyses as a starting point for identifying key and supporting studies to inform the human health hazard assessment, including dose-response analysis. The relevant studies will be evaluated using the data quality criteria in the Application of Systematic Review in TSCA Risk Evaluations document. EPA also expects to consider other studies (e.g., more recently published, alternative test data) that have been published since these reviews, as identified in the literature search conducted by the Agency for perchloroethylene (Perchloroethylene (CASRN 12718-4) Bibliography: Supplemental File for the TSCA Scope Document). EPA expects to consider potential human health hazards associated with perchloroethylene. Based on reasonably available information, the following sections describe the potential hazards associated with perchloroethylene. 2.4.2.1 Non-Cancer Hazards The EPA IRIS Assessment on perchloroethylene (U.S. EPA, 2012e) evaluated the following non-cancer hazards that may be associated with perchloroethylene exposures: the central nervous system (neurotoxicity), kidney, liver and development and reproduction. In general, neurological effects were found to be associated with lower perchloroethylene inhalation exposures. According to the EPA IRIS Assessment (U.S. EPA, 2012e), support for an association with immune and blood effects were less well characterized. In their draft Toxicological Profile for perchloroethylene, ATSDR (2014) identified similar hazard concerns. The National Advisory Committee for Acute Exposure Guideline Levels for Hazardous Substances (NAC/AEGL, 2009) also identified irritation as a hazard concern. WĂŐĞ ϱϭ ŽĨ ϭϲϳ Acute Toxicity Data from acute exposure studies in animals and human incidents indicate that short term exposure to perchloroethylene may cause irritation and neurotoxicity and can impair cognitive function in humans (U.S. EPA, 2012e). An Acute Exposure Guidance Limit (AEGL) values, established by the National Advisory Committee for Acute Exposure Guideline Levels for Hazardous Substances (NAC/AEGL, 2009), has been developed based on irritation to humans (AEGL-1), ataxia in rodents (AEGL-2), and lethality in mice (AEGL-3) (NAC/AEGL, 2009). Neurotoxicity Evidence in humans and animals show that chronic exposure to perchloroethylene can cause neurotoxicity, resulting in decrements in color vision, visuospatial memory and possibly other aspects of cognition and neuropsychological function (U.S. EPA, 2012e). Neurotoxic effects have been characterized in human controlled exposure, occupational exposure and residential studies, as well as in experimental animal studies, providing evidence of an association between perchloroethylene exposure and neurological deficits (U.S. EPA, 2012e). The EPA IRIS assessment for perchloroethylene (U.S. EPA, 2012e) further notes that the nervous system is an expected target with oral perchloroethylene exposures because perchloroethylene and metabolites produced from inhalation exposures will also reach the target tissue via oral exposure. Kidney Toxicity Evidence for kidney toxicity in humans is based on studies of kidney biomarkers, which provide information on nephron integrity and tubule damage. Epidemiologic studies support an association between perchloroethylene and chronic kidney disease (U.S. EPA, 2012e). Animal evidence supports an association between perchloroethylene exposure and chronic kidney disease. Adverse effects on the kidney (e.g., kidney-to-ERG\ ZHLJKW UDWLRV K\DOLQH GURSOHW IRUPDWLRQ JORPHUXODU ³QHSKURVLV ´ NDU\RPHJDO\ (enlarged nuclei), cast formation, and other lesions or indicators of renal toxicity) have been observed in studies of rodents exposed to high concentrations of perchloroethylene by inhalation, oral and intraperitoneal (i.p.) injection of perchloroethylene metabolites (U.S. EPA, 2012e). Liver Toxicity Liver toxicity (i.e., necrosis, vacuolation, etc) has been reported in multiple animal species by inhalation and oral exposures to perchloroethylene, with the mouse typically being more sensitive than the rat (U.S. EPA, 2012e). The liver effects are characterized by increased liver weight, necrosis, inflammatory cell infiltration, triglyceride increases proliferation, cytoplasmic vacuolation (fatty changes), pigment in cells, oval cell hyperplasia and regenerative cellular foci. The EPA IRIS Assessment for perchloroethylene (U.S. EPA, 2012e) found suggestive evidence that perchloroethylene is a liver toxicant in humans. Reproductive/Developmental Toxicity The EPA IRIS Assessment for perchloroethylene (U.S. EPA, 2012e) evaluated the developmental and reproductive toxicity of perchloroethylene in humans and animals. Studies of tetrachloroethylene exposure in humans have evaluated several reproductive outcomes including effects on menstrual disorders, semen quality, fertility, time to pregnancy, and risk of adverse pregnancy outcomes including spontaneous abortion, low birth weight or gestational age, birth anomalies, and stillbirth (U.S. EPA, 2012e). Data from animal studies identified various manifestations of developmental toxicity including, increased mortality and decreased body weight in the offspring of rodents exposed via inhalation. WĂŐĞ ϱϮ ŽĨ ϭϲϳ Irritation U.S. EPA (2012e) and ATSDR (2014) indicate perchloroethylene is irritating. Irritation data for perchloroethylene have also been reviewed outside the EPA IRIS Assessment. Controlled exposures in humans and case reports have identified eye and nose irritation (NAC/AEGL, 2009). 2.4.2.2 Genotoxicity and Cancer Hazards Epidemiologic data provide evidence associating perchloroethylene with several cancer types, including non-Hodgkin lymphoma, multiple myeloma and bladder cancer, with more limited evidence for esophageal, kidney, lung, cervical and breast cancer (U.S. EPA, 2012e). Perchloroethylene is generally considered to be non-genotoxic, however several metabolites exhibit mutagenic and/or genotoxic properties and may contribute to potential genotoxic mode of action (MOA) (U.S. EPA, 2012e). In 2012, EPA released the outcome of the weight-of-evidence cancer assessment, which described the weight-of-evidence judgment of the likelihood that perchloroethylene is a human carcinogen, and quantitative estimates of risk from oral and inhalation exposure (U.S. EPA, 2012e). Following U.S. EPA (2005a) *XLGHOLQHV IRU &DUFLQRJHQ 5LVN $VVHVVPHQW (3$ FRQFOXGHG WKDW SHUFKORURHWK\OHQH LV ³OLNHO\ WR be carcinogenic in humans by DOO URXWHV RI H[SRVXUH´ (U.S. EPA, 2012e). 2.4.2.3 Potentially Exposed or Susceptible Subpopulations TSCA requires that the determination of whether a chemical substance presents an unreasonable risk include consideration of unreasonable risk WR ³D SRWHQWLDOO\ H[SRVHG RU VXVFHSWLEOH VXESRSXODWLRQ LGHQWLILHG DV UHOHYDQW WR WKH ULVN HYDOXDWLRQ´ E\ (3$ 76&$ † VWDWHV WKDW ³WKH WHUP µSRWHQWLDOO\ H[SRVHG RU VXVFHSWLEOH VXESRSXODWLRQ¶ PHDQV D JURXS RI LQGLYLGXDOV ZLWKLQ WKH JHQHUDO SRSXODWLRQ identified by the Administrator who, due to either greater susceptibility or greater exposure, may be at greater risk than the general population of adverse health effects from exposure to a chemical substance or mixture, such as infants, children, pregnant ZRPHQ ZRUNHUV RU WKH HOGHUO\ ´ In developing the hazard assessment, EPA will analyze available data to ascertain whether some human receptor groups may show greater susceptibility WR WKH FKHPLFDO¶V KD]DUGV GXH WR LQWULQVLF IDFWRUV (3$ plans to analyze the susceptibility factors identified in the EPA IRIS assessment for perchloroethylene U.S. EPA (2012e) and ATSDR (2014) evaluations. These assessments both identified the following subpopulations as possibly more susceptible to adverse effects associated with perchloroethylene exposures: early and later lifestages and groups defined by health and nutrition status, gender, race/ethnicity, genetics and multiple exposures and cumulative risk. However U.S. EPA (2012e) also determined that the available data was insufficient to allow for a quantitative assessment of the impact of susceptibility on risk. The California Office of Environmental Health Hazard Assessment OEHHA (2016) derived an inhalation cancer unit risk factor for perchloroethylene based on the same physiologically based pharmacokinetic (PBPK) model (Chiu and Ginsberg, 2011) used in the EPA IRIS assessment (U.S. EPA, 2012e). The model included both oxidative metabolism and glutathione conjugation metabolism; the latter varies greatly within the human population, with some variation representing sensitive subpopulations (Spearow et al., 2017; OEHHA, 2016). EPA will consider this information during the risk evaluation phase. 2.5 Conceptual Models EPA risk assessment guidance (U.S. EPA, 2014d), defines Problem Formulation as the part of the risk assessment framework that identifies the major factors to be considered in the assessment. It draws from the regulatory, decision-PDNLQJ DQG SROLF\ FRQWH[W RI WKH DVVHVVPHQW DQG LQIRUPV WKH DVVHVVPHQW¶V technical approach. WĂŐĞ ϱϯ ŽĨ ϭϲϳ A conceptual model describes the actual or predicted relationships between the chemical substance and receptors, either human or environmental. These conceptual models are integrated depictions of the conditions of use, exposures (pathways and routes), hazards and receptors. The initial conceptual models describing the scope of the assessment for perchloroethylene, have been refined during problem formulation. The changes to the conceptual models in this problem formulation are described along with the rationales. In this section EPA outlines those pathways that will be included and further analyzed in the TSCA risk evaluation; will be included but will not be further analyzed in risk evaluation; and will not be included in the TSCA risk evaluation and the underlying rationale for these decisions. EPA determined as part of problem formulation that it is not necessary to conduct further analysis on certain exposure pathways that were identified in the perchloroethylene scope document and that remain in the risk evaluation. Each risk evaluation will be "fit-for-purpose," meaning not all conditions of use will warrant the same level of evaluation and the Agency may be able to reach some conclusions without extensive or quantitative risk evaluations. 82 FR 33726, 33734, 33739 (July 20, 2017). As part of this problem formulation, EPA also identified exposure pathways under regulatory programs of other environmental statutes, administered by EPA, which adequately assess and effectively manage exposures and for which long-standing regulatory and analytical processes already exist, i.e., the Clean Air Act (CAA), the Safe Drinking Water Act (SDWA), the Clean Water Act (CWA) and the Resource Conservation and Recovery Act (RCRA). OPPT worked closely with the offices within EPA that administer and implement the regulatory programs under these statutes. In some cases, EPA has determined that chemicals present in various media pathways (i.e., air, water, land) fall under the jurisdiction of existing regulatory programs and associated analytical processes carried out under other EPA-administered statutes and have been assessed and effectively managed under those programs. EPA believes that the TSCA risk evaluation should generally focus on those exposure pathways associated with TSCA conditions of use that are not adequately assessed and effectively managed under the regulatory regimes discussed above because these pathways are likely to represent the greatest areas of risk concern. As a result, EPA does not expect to include in the risk evaluation certain exposure pathways identified in the perchloroethylene scope document. Conceptual Model for Industrial and Commercial Activities and Uses: Potential Exposures and Hazards The revised conceptual model (Figure 2-2) describes the pathways of exposure from industrial and commercial activities and uses of perchloroethylene that EPA expects to include in the risk evaluation. There are exposures to workers and/or occupational non-users via inhalation routes and/or exposures to workers via dermal routes for all conditions of use identified in this problem formulation. In addition to the pathways illustrated in the figure, EPA will evaluate activities resulting in exposures associated with distribution in commerce (e.g. loading, unloading) throughout the various lifecycle stages and conditions of use (e.g. manufacturing, processing, industrial use, commercial use, disposal) rather than a single distribution scenario. Inhalation Inhalation exposures for workers DUH UHJXODWHG E\ 26+$¶V RFFXSDWLRQDO VDIHW\ DQG KHDOWK VWDQGDUGV IRU perchloroethylene which include a PEL of 100 ppm TWA, exposure monitoring, control measures and respiratory protection (29 CFR 1910.134). EPA expects that for workers and occupational non-users exposure via inhalation will be the most significant route of exposure for most exposure scenarios. EPA WĂŐĞ ϱϰ ŽĨ ϭϲϳ Conceptual Model for Environmental Releases and Wastes: Potential Exposures and Hazards The revised conceptual model (Figure 2-4) illustrates the expected exposure pathways to human (i.e., general population) and ecological receptors (i.e., aquatic and terrestrial) from environmental releases and waste streams associated with industrial and commercial activities for perchloroethylene that EPA expects to include in the risk evaluation. The pathways that EPA expects to include and analyze further in the risk evaluation is described in Section 2.5.3.1 and shown in the conceptual model Figure 2-4. The pathways that EPA does not expect to include in the risk evaluation s are described in Section 2.5.3.2. 2.5.3.1 Pathways That EPA Expects to Include and Further Analyze in the Risk Evaluation EPA plans to analyze aquatic organisms exposed via contaminated surface water. There are no national recommended water quality criteria for the protection of aquatic life for perchloroethylene and as a result EPA does not believe that perchloroethylene exposure to aquatic organisms in surface water has been adequately assessed or effectively managed under other EPA statutory authorities (see Section 2.5.3.2). EPA identified and reviewed national scale monitoring data to support this problem formulation. EPA and the USGS National Water Quality Assessment Program (Cycle 1, 1992-2001) reported perchloroethylene contamination in U.S. surface water and ground water in 19.6% of samples (n=5,911) and at 13.2% of sites (n=4,295), with detection in surface water occurring more frequently than in ground water (U.S. EPA, 2009). More recently measured, nationalVFDOH PRQLWRULQJ GDWD ZDV IURP (3$¶V 672UDJH DQG 5(7UHLYDO 6725(7 DQG 1DWLRQDO :DWHU Information System (NWIS). Based on STORET query for perchloroethylene for the past ten years, perchloroethylene is detected in surface water in the United States. The data showed a detection rate (above quantification limit and/or above reporting limit) of approximately 15% for surface water, with detections ranging from 0.02 —g/L to 26.7 —g/L. As summarized in Section 2.4.1 perchloroethylene showed hazard at concentrations as low as 14 —g/L for aquatic plants. The chronic COC value of 1 —g/L is not sufficiently below the range of monitored concentrations to eliminate risk concerns. Therefore, EPA plans to evaluate risks to aquatic organisms from exposures to perchloroethylene in surface waters. 2.5.3.2 Pathways That EPA Does Not Expect to Include in the Risk Evaluation Exposures to receptors may occur from industrial and/or commercial uses, industrial releases to air, water or land; and other conditions of use. As described in section 2.5, pathways under other environmental statutes, administered by EPA, which adequately assess and effectively manage exposures and for which long-standing regulatory and analytical processes already exist will not be included in the risk evaluation. These pathways are described below. Ambient Air Pathway The Clean Air Act (CAA) contains a list of hazardous air pollutants (HAP) and provides EPA with the authority to add to that list pollutants that present, or may present, a threat of adverse human health effects or adverse environmental effects. For stationary source categories emitting HAP, the CAA requires issuance of technology-based standards and, if necessary, additions or revisions to address developments in practices, processes, and control technologies, and to ensure the standards adequately protect public health and the environment. The CAA thereby provides EPA with comprehensive authority to regulate emissions to ambient air of any hazardous air pollutant. WĂŐĞ ϱϵ ŽĨ ϭϲϳ Perchloroethylene is a HAP. EPA has issued a number of technology-based standards for source categories that emit perchloroethylene to ambient air and, as appropriate, has reviewed, or is in the process of reviewing remaining risks. Because stationary source releases of perchloroethylene to ambient air are adequately assessed and any risks effectively managed when under the jurisdiction of the CAA, EPA does not plan to evaluate emission pathways to ambient air from commercial and industrial stationary sources or associated inhalation exposure of the general population or terrestrial species in this TSCA evaluation. Drinking Water Pathway EPA has regular analytical processes to identify and evaluate drinking water contaminants of potential regulatory concern for public water systems under the Safe Drinking Water Act (SDWA). Under 6':$ (3$ PXVW DOVR UHYLHZ DQG UHYLVH ³DV DSSURSULDWH´ H[LVWLng drinking water regulations every 6 years. EPA has promulgated National Primary Drinking Water Regulations (NPDWRs) for perchloroethylene under the Safe Drinking Water Act. EPA has set an enforceable Maximum Contaminant Level (MCL) as close as feasible to a health based, non-enforceable Maximum Contaminant Level Goal (MCLG). Feasibility refers to both the ability to treat water to meet the MCL and the ability to monitor water quality at the MCL, SDWA Section 1412(b)(4)(D), and public water systems are required to monitor for the regulated chemical based on a standardized monitoring schedule to ensure compliance with the (MCL). Hence, because the drinking water exposure pathway for perchloroethylene is currently addressed in the SDWA regulatory analytical process for public water systems, EPA does not plan to include this pathway LQ WKH ULVN HYDOXDWLRQ IRU SHUFKORURHWK\OHQH XQGHU 76&$ LQ WKH ULVN HYDOXDWLRQ IRU SHUFKORURHWK\OHQH XQGHU 76&$ (3$¶V 2IILFH RI :DWHU DQG 2IILFH RI Pollution Prevention and Toxics will continue to work together providing understanding and analysis of the SDWA regulatory analytical processes and to exchange information related to toxicity and occurrence data on chemicals undergoing risk evaluation under TSCA. Ambient Water Pathways EPA develops recommended water quality criteria under section 304(a) of the CWA for pollutants in surface water that are protective of aquatic life or human health designated uses. EPA develops and publishes water quality criteria based on priorities of states and others that reflect the latest scientific knoZOHGJH $ VXEVHW RI WKHVH FKHPLFDOV DUH LGHQWLILHG DV ³SULRULW\ SROOXWDQWV´ KXPDQ KHDOWK DQG aquatic life). The CWA requires states adopt numeric criteria for priority pollutants for which EPA has published recommended criteria under section 304(a), the discharge or presence of which in the affected waters could reasonably be expected to interfere with designated uses adopted the state. When states DGRSW FULWHULD WKDW (3$ DSSURYHV DV SDUW RI VWDWH¶V UHJXODWRU\ ZDWHU TXDOLW\ VWDQGDUGV H[SRVXUH LV considered when state permit writers determine if permit limits are needed and at what level for a specific discharger of a pollutant to ensure protection of the designated uses of the receiving water. Once states adopt criteria as water quality standards, the CWA requires National Pollutant Discharge Elimination System (NPDES) discharge permits include effluent limits as stringent as necessary to meet standards. CWA section 301(b)(1)(C). This is the process used under the CWA to address risk to human health and aquatic life from exposure to a pollutant in ambient waters. EPA has identified perchloroethylene as a priority pollutant and EPA has developed recommended water quality criteria for protection of human health for perchloroethylene which are available for adoption into state water quality standards for the protection of human health and are available for use by NPDES permitting authorities in deriving effluent limits to meet state narrative criteria. As such, WĂŐĞ ϲϬ ŽĨ ϭϲϳ EPA does not expect to include this pathway LQ WKH ULVN HYDOXDWLRQ XQGHU 76&$ LQ WKH ULVN HYDOXDWLRQ XQGHU 76&$ (3$¶V 2IILFH RI :DWHU and Office of Pollution Prevention and Toxics will continue to work together providing understanding and analysis of the CWA water quality criteria development process and to exchange information related to toxicity of chemicals undergoing risk evaluation under TSCA. EPA may update its CWA section 304(a) water quality criteria for perchloroethylene in the future under the CWA. EPA has not developed CWA section 304(a) recommended water quality criteria for the protection of aquatic life for perchloroethylene, so there are no national recommended criteria for this use available for adoption into state water quality standards and available for use in NPDES permits. As a result, this pathway will undergo aquatic life risk evaluation under TSCA (see Section 2.4.1). EPA may publish CWA section 304(a) aquatic life criteria for perchloroethylene in the future if it is identified as a priority under the CWA. Biosolids Pathways CWA Section 405(d) requires EPA to 1) promulgate regulations that establish numeric criteria and management practices that are adequate to protect public health and the environment from any reasonably anticipated adverse effects of toxic pollutants during the use or disposal of sewage sludge, and 2) review such regulations at least every two years to identify additional toxic pollutants that occur LQ ELRVROLGV L H ³%LHQQLDO 5HYLHZV´ DQG UHJXODWH WKRVH SROOXWDQWV LI VXIILFLHQW VFLHQWLILF HYLGHQFH shows they may be present in sewage sludge in concentrations which may adversely affect public health or the environment. EPA also periodically conducts surveys to determine what may be present in sewage sludge. EPA has conducted four sewage sludge surveys and identified compounds that occur in biosolids in seven Biennial Reviews. EPA has regulated 10 chemicals in biosolids under CWA 405(d). EPA has identified perchloroethylene in biosolids biennial reviews. The purpose of such reviews is to identify additional toxic pollutants in biosolids. EPA can potentially regulate those pollutants under &:$ G EDVHG RQ D VXEVHTXHQW DVVHVVPHQW RI ULVN (3$¶V 2IILFH RI :DWHU LV FXUUHQWO\ GHYHORSLQJ modeling tools in order to conduct risk assessments for chemicals in biosolids. Because the biosolids pathway for perchloroethylene is currently being addressed in the CWA regulatory analytical process, this pathway will not be further analyzed in the risk evaluation for perchloroethylene under TSCA. (3$¶V 2IILFH RI :DWHU DQG 2IILFH RI 3ROOXWLRQ 3UHYHQWLRQ DQd Toxics will continue to work together to discuss significant data gaps and exchange information related to exposure and toxicity of this chemical as OW conducts the risk assessment under the CWA. Disposal Pathways Perchloroethylene is included on the list of hazardous wastes pursuant to RCRA 3001 (40 CFR §§ 261.33) as a listed waste on the F, K and U lists. The general RCRA standard in Section RCRA 3004(a) for the technical criteria that govern the management (treatment, storage, and disposal) of hazardous waste are those "necessary to protect human health and the environment," RCRA 3004(a). The UHJXODWRU\ FULWHULD IRU LGHQWLI\LQJ ³FKDUDFWHULVWLF´ KD]DUGRXV ZDVWHV DQG IRU ³OLVWLQJ´ D ZDVWH DV hazardous also relate solely to the potential risks to human health or the environment. 40 C.F.R. §§ 261.11, 261.21- 5&5$ VWDWXWRU\ FULWHULD IRU LGHQWLI\LQJ KD]DUGRXV ZDVWHV UHTXLUH (3$ WR ³WDN>H@ into account toxicity, persistence, and degradability in nature, potential for accumulation in tissue, and other UHODWHG IDFWRUV VXFK DV IODPPDELOLW\ FRUURVLYHQHVV DQG RWKHU KD]DUGRXV FKDUDFWHULVWLFV ´ Subtitle C control cover not only hazardous wastes that are landfilled, but also hazardous wastes that are incinerated (subject to joint control under RCRA Subtitle C and the Clean Air Act (CAA) hazardous waste combustion MACT) or injected into UIC Class I hazardous waste wells (subject to joint control under Subtitle C and the Safe Drinking Water Act (SDWA)). WĂŐĞ ϲϭ ŽĨ ϭϲϳ EPA does not expect to include emissions to ambient air from municipal and industrial waste incineration and energy recovery units in the risk evaluation, as they are regulated under section 129 of the Clean Air Act. CAA section 129 also requires EPA to review and, if necessary, add provisions to ensure the standards adequately protect public health and the environment. Thus, combustion byproducts from incineration treatment of perchloroethylene wastes (the majority of the 1.1 million lbs identified as treated in Tables 2-6 ± 2-8) would be subject to these regulations, as would perchloroethylene burned for energy recovery (2.3 million lbs). EPA does not expect to include on-site releases to land that go to underground injection in its risk evaluation. TRI reporting in 2016 indicated 272 pounds released to underground injection to a Class I well and no releases to underground injection wells of Classes II-VI. Environmental disposal of perchloroethylene injected into Class I well types managed and prevented from further environmental release by RCRA and SDWA regulations. Therefore, disposal of perchloroethylene via underground injection is not likely to result in environmental and general population exposures. EPA does not expect to include on-site releases to land from RCRA Subtitle C hazardous waste landfills or exposures of the general population (including susceptible populations) or terrestrial species from such releases in the TSCA evaluation. Based on 2015 reporting to TRI, the majority of the land disposals occur in Subtitle C landfills (78,120 lbs). Design standards for Subtitle C landfills require double liner, double leachate collection and removal systems, leak detection system, run on, runoff, and wind dispersal controls, and a construction quality assurance program. They are also subject to closure and post-closure care requirements including installing and maintaining a final cover, continuing operation of the leachate collection and removal system until leachate is no longer detected, maintaining and monitoring the leak detection and groundwater monitoring system. Bulk liquids may not be disposed in Subtitle C landfills. Subtitle C landfill operators are required to implement an analysis and testing program to ensure adequate knowledge of waste being managed, and to train personnel on routine and emergency operations at the facility. Hazardous waste being disposed in Subtitle C landfills must also meet RCRA waste treatment standards before disposal. Given these controls, general population exposure to perchloroethylene in groundwater from Subtitle C landfill leachate is not expected to be a significant pathway. EPA does not expect to include on-site releases to land from RCRA Subtitle D municipal solid waste landfills or exposures of the general population (including susceptible populations) or terrestrial species from such releases in the TSCA evaluation. While permitted and managed by the individual states, municipal solid waste (MSW) landfills are required by federal regulations to implement some of the same requirements as Subtitle C landfills. MSW landfills generally must have a liner system with leachate collection and conduct groundwater monitoring and corrective action when releases are detected. MSW landfills are also subject to closure and post-closure care requirements, and must have financial assurance for funding of any needed corrective actions. MSW landfills have also been designed to allow for the small amounts of hazardous waste generated by households and very small quantity waste generators (less than 220 lbs per month). Bulk liquids may not be disposed in Subtitle C landfills. EPA does not expect to include on-site releases to land from industrial non-hazardous waste and construction/demolition waste landfills in the perchloroethylene risk evaluation. Industrial nonhazardous and construction/demolition waste landfills are primarily regulated under state regulatory programs. S States must also implement limited federal regulatory requirements for siting, groundwater monitoring and corrective action and a prohibition on open dumping and disposal of bulk liquids. States WĂŐĞ ϲϮ ŽĨ ϭϲϳ may also establish additional requirements such as for liners, post-closure and financial assurance, but are not required to do so. Therefore, EPA does not expect to include this pathway in the risk evaluation. WĂŐĞ ϲϯ ŽĨ ϭϲϳ Testimony of Kimberly W. White, Ph.D. Senior Director, Chemical Products and Technology Division American Chemistry Council Making EPA Great Again House Committee on Science, Space, and Technology February 7, 2017 americanchemistry.com(R) 700 Second St., NE | Washington, DC 20002 | (202) 249.7000 Summary The American Chemistry Council (ACC)1 appreciates this opportunity to provide testimony on the U.S. Environmental Protection Agency's (EPA) process for evaluating and using science to support regulatory decision making. The business of chemistry is a critical component for manufacturing safe, high quality products and ACC member companies rely on science to conduct the research necessary to discover new chemistries and identify new applications of existing chemistries. They also rely on science to develop new tools for assessing the potential hazards, exposures and risks of chemical substances. Similarly, they expect high quality, up to date science and relevant reliable assessment processes to underpin regulatory decisions by the Federal government. Reliance on the highest quality, best available science is critical to ensuring public trust. Without it, consumers are at a severe disservice and lose confidence in regulatory decision making, leading to product de-selection that is not supported by science, unwarranted public alarm and unnecessary costs. ACC supports actions to enhance the integration of the best available scientific knowledge and methods as the foundation for regulatory decision making across EPA. We also support advancing the technical quality and objectivity of EPA evaluations, particularly through enhancing transparency in both what science is being considered and how it is being interpreted and integrated. Over the last 30 years, advances in scientific techniques and knowledge have improved our understanding of how chemicals interact with the human body and the environment. Research programs within industry, academia and government have expanded to investigate the underlying biological processes for chemical interactions and to improve the scientific basis of chemical policies and product stewardship efforts. It is simply not enough to have the science available for use. There must also be a transparent process and a willingness to enable integration of the science into EPA policies and practices. Current processes for evaluating scientific information and conducting chemical assessments at EPA are not always based on transparent, objective or consistent use of the best available science. In recent years, there has been a focus on EPA's Integrated Risk Information System (IRIS) program and addressing deficiencies identified by the National Academy of Sciences (NAS)2. These deficiencies are also evident in other EPA chemical assessment programs. 1 The American Chemistry Council (ACC) represents the leading companies engaged in the business of chemistry. ACC members apply the science of chemistry to make innovative products and services that make people's lives better, healthier and safer. ACC is committed to improved environmental, health and safety performance through Responsible Care(R), common sense advocacy designed to address major public policy issues, and health and environmental research and product testing. The business of chemistry is a $797 billion enterprise and a key element of the nation's economy. It is one of the nation's largest exporters, accounting for fourteen percent of all U.S. exports. Chemistry companies are among the largest investors in research and development. 2 National Academy of Sciences (NAS). NRC (National Research Council). Review of EPA's Integrated Risk Information System (IRIS) Process. Board on Environmental Studies and Toxicology. Division on Earth and Life Studies. Washington, DC: The National Academies Press, 2014. Available at http://www.nap.edu/catalog.php?record_id=18764. 2 ACC has consistently called upon EPA to improve the design and conduct of its chemical assessments. Recommended improvements have included adoption of consistent and transparent study evaluation methods to determine the quality and reliability of critical studies. We have also encouraged EPA to utilize an improved framework for integrating study results based on a weight of the scientific evidence approach that incorporates modern knowledge of mode of action to establish cause and effect. Furthermore, we have recommended that EPA improve its peer review and accountability practices for addressing both public comments and peer review recommendations. Although EPA has made efforts to identify practices for systematically reviewing the available science and to strengthen its peer review processes, the actual implementation of these practices has been slow and often lacking. This has been fundamentally due to the lack of a consistent science-based framework for conducting chemical evaluations within EPA. Over the past several years Congress has worked to update and reform the Toxic Substances Control Act (TSCA) and in 2016 passed the Frank R. Lautenberg Chemical Safety for the 21st Century Act (LCSA). The LCSA establishes new requirements for the review of new and existing chemicals manufactured and used to make U.S. products; including requiring use of the best available science and a weight of evidence process to evaluate scientific information. EPA now has a mandate to apply high quality, reliable scientific information while evaluating new chemicals and prioritizing and evaluating the risks of existing chemicals. Implementing these new provisions will require significant changes to EPA's scientific evaluation procedures, particularly for existing chemicals. However, as we have recently seen in EPA's proposed framework rule for risk evaluation, EPA believes that existing practices meet the standards of the LCSA. ACC does not support this belief and we plan to continue to be a constructive partner to both Congress and EPA in identifying approaches that enhance the chemical assessment process. ACC's testimony today outlines four areas that can improve the evaluation of scientific information at EPA: ? ? ? ? I. Clear framework for conducting the chemical assessment; Application of consistent criteria for selecting and evaluating scientific data; Transparent and objective integration of scientific evidence; and Independent and robust peer review. Clear Framework for Conducting Chemical Assessment EPA and other federal government agencies conduct chemical assessments to inform risk management decisions. As such, EPA should ensure that the assessments it conducts will address the information needs of decision makers. EPA is tasked with evaluating new chemicals to be manufactured and used to make U.S. products and existing chemicals in the marketplace. As such, any assessment that EPA undertakes should be fit for purpose in order to effectively and efficiently utilize its limited resources. This can help ensure that chemical assessments are based on the best available information and are appropriately scaled and oriented to address relevant questions regarding risk. EPA should also make use of all available science evaluations, including primary scientific literature, grey literatures and reviews, conducted to inform the chemical assessment process and determine information needs. In this initial phase of chemical assessment, EPA can determine if a screening level 3 assessment will identify and assess risk sufficiently or if a more refined risk evaluation is needed. EPA is currently interpreting the LCSA as requiring the Agency to evaluate all conditions of use of a chemical, regardless of how small, in the risk evaluation. This interpretation is unreasonable and inconsistent with other provisions in the LCSA which, clearly indicate that EPA has discretion to select the conditions of use that it will consider in the scope of a risk evaluation. There are significant questions about EPA's ability to meet the stringent evaluation deadlines of the LCSA if the Agency takes the position that all uses of a substance must be evaluated. A tiered approach, where EPA uses the scoping step to conduct a quantitative screening level analysis, will allow EPA to focus its limited resources on more robust refined risk evaluations for only those conditions of use where unreasonable risks cannot be ruled out. In order to adequately and effectively evaluate the available science to make timely and science based decisions regarding potential risk from exposures, methods for conducting a chemical assessment must be clearly defined up front. The protocol, developed before the chemical evaluation begins, defines the methodologies that will be used in the assessment. It is made publicly available before the assessment begins and becomes a living document that can be commented upon and modified as needed. The protocol includes: a clear testable question/hypothesis, the planned search strategy (including criteria for inclusion and exclusion of studies), the criteria that will be used for study quality and risk of bias evaluations (including, for example, consideration of study design and confounders), the plan for integrating/synthesizing scientific evidence using a weight of evidence approach, the plan for quantifying and presenting findings, and the plan for peer review of the assessment. Section 6(b)(4)(B) of the LCSA requires EPA to establish, by rule, "a process to conduct risk evaluations." Incorporation of a protocol which includes these important risk evaluation elements is missing from EPA's proposed rule for risk evaluation. Without these elements it is not clear how EPA can meet the LCSA requirements that, for the first time in federal law, provide a statutory requirement mandating best available science and weight of the scientific evidence requirements to inform agency decision making. In EPA's IRIS program there are similar concerns regarding scientific evaluation procedures. These concerns have been well articulated by the NAS. Assessment approaches also appear to be hampered by a lack of coordination among programs regarding the chemical assessments being undertaken and how those assessments can be utilized by other EPA programs. For instance, past assessments by EPA's IRIS program (e.g., n-butanol, trimethylbenzenes) did not seem to consider data developed by other EPA program offices in previous chemical assessments. It also has not been clear why the TSCA Work Plan chemicals program, within the Office of Pollution Prevention and Toxics (OPPT), at times evaluated the same chemicals that the IRIS program evaluated. ACC recommends that EPA identify a consistent framework for conducting chemical assessments, including the methods to be utilized in the assessment and the utility of the assessment for regulatory decision making, prior to initiating the assessment These practices should be consistent with requirements outlined in section 26 of the LCSA which, requires 4 EPA to improve the quality, transparency and relevance of the scientific information, approaches, methods, protocols, and models that are used to evaluate chemical risks. EPA must additionally ensure that the information used is reasonable for and consistent with the intended use of the information.3 When assessments are being conducted to inform significant rulemakings, EPA must make certain that these important standards are being met. II. Application of Consistent Criteria for Selecting and Evaluating Scientific Data Scientific evaluations must utilize transparent and consistent criteria for selecting the most relevant scientific information and evaluating the evidence to draw scientifically defensible conclusions to support decision making. In particular, a systematic approach can ensure that EPA is using clear procedures and protocols to develop reproducible and scientifically sound assessments. It is critical that EPA rely on the studies of the highest quality not simply those studies that produce the lowest points of departure or the highest exposure estimates. A lack of sufficient review of study information may lead to establishing unrealistic risk characterizations and exposure concentrations that are not relevant to actual human exposures. For example, in the Work Plan chemical draft risk assessment of 1bromopropane, EPA did not provide information regarding the quality of the individual studies. Appendix M of the assessment identifies some quality considerations, but EPA did not provide any information regarding its own findings from its quality review of the individual studies. Additionally, no information was provided to describe how considerations were applied and what constitutes a study of "high quality" or "good quality."4 Simply choosing the lowest value is not consistent with the best available science approach envisioned under the LCSA. As noted before, this new language will require that EPA make significant changes to its risk evaluation practices. Given the lack of consistency in evaluating scientific information, EPA should develop, through an open and transparent process, (1) protocols that define the process for the acquisition of the scientific literature including study inclusion/exclusion criteria; and (2) a framework for evaluating studies for quality, reliability and relevance. Notably, the LCSA calls on the EPA risk evaluation process to comply with the best available science provision in Section 26(h), the weight of the scientific evidence provision in 26(i) and the transparency provision in 26(j). III. Transparent and Objective Integration of Scientific Evidence Considerable progress has been made over the years to improve understanding of the potential for risk from chemical exposures. In order for the Agency to reach scientifically robust conclusions, it must employ a transparent weight of evidence framework that 3 Section 26(h)(1) states that the Administrator shall consider "the extent to which the scientific information, technical procedures, measures, methods, protocols, methodologies, or models employed to generate the information are reasonable for and consistent with the intended use of the information." 4 See Comments of the American Chemistry Council on the TSCA Work Plan Chemical Draft Risk Assessment of 1-Bromopropane , Docket No. EPA-HQ-OPPT-2015-0084, May 9, 2016 5 integrates evidence from human epidemiological studies, laboratory animal research and mechanistic research. This includes evaluating the strengths and limitations of the human and animal data, understanding the biological significance of responses in animal models and of mechanistic research, and applying current scientific knowledge to extrapolate those findings to humans. EPA's 2005 Guidelines for Carcinogen Risk Assessment5 emphasize the importance of weighing all of the evidence--including both studies that provide evidence of an effect as well as those that provide no evidence of an effect--in reaching conclusions about the potential for a chemical to be carcinogenic to humans. The weighing of the evidence includes addressing not only the likelihood of human carcinogenicity, but also the conditions under which such effects might occur. Weighing the scientific evidence entails clearly explaining the kinds of evidence available (e.g., epidemiology, toxicology, mechanistic) and how that evidence fits together in drawing conclusions regarding human relevance and dose-response. Section 6(b)(4)(F)(i) of the LCSA requires risk evaluations to integrate and assess available information on hazards and exposures for the conditions of use of the chemical substance. Additionally, Section 26(i) of the LCSA requires EPA to make decisions using a weight of scientific evidence approach. The Congressional Record clearly describes how a weight of the scientific evidence approach requires the consideration of the strengths, limitations and relevance of each study.6 Unfortunately, it has been unclear how the EPA programs apply weight of evidence approaches or how the programs incorporate mode of action information when evaluating the science to reach decisions. There also appears a lack of acknowledgement in some EPA programs regarding science that supports a threshold for safe exposures to a chemical. In 2011, the NAS7 reviewed the draft formaldehyde IRIS assessment and concluded that EPA had not sufficiently documented methods to identify or evaluate relevant scientific studies; and had not adequately integrated the lines of evidence from the available animal, human, and mechanistic data. The NAS report also called the draft formaldehyde assessment subjective and potentially problematic given the inconsistencies in the available scientific data. The NAS also noted areas where EPA's approaches may not be scientifically justified. For example, the NAS review noted that with regard to the biologically based dose-response (BBDR) model manipulations made by the EPA "...some of the manipulations are extreme, may not be scientifically justified, and should not have been used as the basis of rejection of the use of the BBDR model in its assessment. Model manipulations that yield results that are implausible or inconsistent with available data should be rejected as a basis for judging the utility of the model." 5 EPA 2005. Guidelines for Carcinogen Risk Assessment. Available at http://www.epa.gov/sites/production/files/2013-09/documents/cancer_guidelines_final_3-25-05.pdf. 6 See Senate Congressional Record, June 7, 2016 at page S3518, available at: https://www.congress.gov/crec/2016/06/07/CREC-2016-06-07-pt1-PgS3511.pdf. 7 National Academy of Sciences (NAS). National Research Council (NRC). Review of the Environmental Protection Agency's Draft IRIS Assessment of Formaldehyde. Washington, DC: The National Academies Press, 2011. 6 In addition to identifying scientific concerns with the formaldehyde IRIS assessment the NAS also identified recurrent problems with EPA's process for evaluating chemicals more broadly. While the EPA is working to address these NAS recommendations, after more than 5 years, the IRIS program is still falling short and has not yet released a final assessment that is fully consistent with these important NAS recommendations. In addition to the IRIS program, and more recently, in a 2016 Work Plan chemical review of 1-bromopropane, EPA had multiple studies for identified hazards, such as reproductive and developmental toxicity, and carcinogenicity. EPA also had multiple exposure studies to consider. However, the Agency failed to apply a weight of evidence approach. When there are multiple studies available, the only scientifically defensible approach is to weigh the studies by considering study characteristics and determining which studies are of higher quality and should be given greater weight in the assessment. Failure to employ a weight of evidence approach is a critical deficiency that seriously limits any conclusions that can be drawn. To ensure clarity and consistency in the scientific evaluation process, EPA should (1) develop a clear weight of evidence framework to identify and evaluate each stream of evidence, including strengths, limitations, and relevance of each study; and (2) integrate evidence based upon strengths, limitations, and relevance. This approach should be implemented in all programs and codified in EPA's risk evaluation framework regulations under the LCSA. IV. Independent and Robust Peer Review Peer review is essential in the evaluation of scientific information to ensure the development of scientifically defensible assessments. It also allows for the review of the underlying assumptions, methodology, criteria, and conclusions reached in the evaluation. EPA currently has several mechanisms to conduct peer review of scientific information including the Science Advisory Board, Science Advisory Panel, NAS contracted review and ad hoc peer review. As outlined in EPA's 2015 Peer Review Handbook,8 "the success and usefulness of any peer review depends on the quality of the draft work product submitted for peer review, the care given to the statement of the issues or "charge," the match between the peer review draft product and the form of peer review, the match between the peer review draft product and the scientific/technical expertise of the reviewers, and Agency use of peer review comments in the final product." Unfortunately, peer review processes and approaches are inconsistently applied throughout the Agency, including the selection of peer review panel members and the consideration given to public and peer review comments. During some EPA peer review meetings, the peer reviewers have appeared to be overly deferential to EPA and reluctant to be seen as criticizing EPA staff. We have also seen situations where peer reviewers have suggested discounting a study solely based on the funding source, without any considerations being given to the quality of the study. Also, EPA staff often comment throughout peer review meetings, essentially participating as peers, while industry experts are typically excluded 8 EPA Peer Review Handbook 4th Edition, 2015. Available at https://www.epa.gov/sites/production/files/201603/documents/epa_peer_review_handbook_4th_edition.pdf 7 from the dialogue. This practice undermines the integrity of the reviewers' role as independent and external to the assessment itself. A critical element of peer review is also the consideration of public comments. The public plays an important role in the review process by helping identify key scientific information and potential concerns with the assessment being evaluated. Currently, there is no robust consideration of public comments in the peer review process. Reviewers on the EPA Science Advisory Board (SAB) are not given clear advice regarding what it means to "consider" public comments. In fact we have seen SAB chairs ignore public input because they are not required to address it. When this has occurred, SAB staff have not clarified to the peer reviewers that they can and should respond to public input. In 2013, EPA's IRIS program announced a revised process that included an explicit response to comments step. However, 2016 IRIS assessments of trimethylbenzenes and ammonia and the 2017 ethylene oxide assessment contained no response to public comments in the final documents and only addressed peer review comments. This is a clear departure from EPA's commitment in step 5 of its IRIS process which states that the Agency "Develops a disposition of peer review and public comments and provides these as an appendix to the IRIS assessment."9 Compounding concerns, the SAB committee reviewing the trimethylbezene assessment also did not respond to public comments, essentially creating a black hole where public comments are provided to the Agency but no clear responses are provided. Peer review should be independent and objective allowing for robust engagement with stakeholders to provide a thorough review. It should also include a quality assurance process that explicitly evaluates whether the peer review recommendations and public comments were completely and adequately addressed. Conclusion The incorporation of up to date scientific information, approaches and methods to ensure that EPA decision making is firmly based on high quality science is critical to protecting human health and the environment. This can be achieved by transparent, objective and consistent application of evaluation processes throughout EPA to evaluate and integrate scientific information utilizing a weight of scientific evidence process as required under the LCSA. Further, a robust and independent peer review process must be employed. ACC looks forward to working with members of the Committee to enhance the approaches to ensure that high quality science is the foundation to regulatory decision making regarding potential chemical hazards and risks. 9 EPA IRIS Process Flow Chart. Available at https://www.epa.gov/sites/production/files/201403/documents/iris_process_flow_chart.pdf. 8 Kimberly W. White is a Senior Director in the American Chemistry Council's Chemical Products and Technology Division where she works in support of scientific research and chemical assessments that are firmly based on up-to-date scientific knowledge, meet the highest standards of scientific inquiry and are evaluated in accordance with the most relevant scientific approaches. For the past 5 years, Dr. White has served as a scientific advisor to industry partners for the development and execution of scientific research to inform chemical hazard assessments. She has also worked to identify emerging issues and trends in science policy and risk evaluation. Dr. White has presented at scientific symposia; collaborated to organize multi-stakeholder workshops to improve the conduct of chemical assessments; and managed scientific research programs. Additionally, Dr. White has coauthored publications on weight of evidence frameworks, problem formulation in chemical assessment and understanding potency information associated with human exposures. She has also been the lead representative in discussions with regulatory and chemical assessment agencies. In her most recent past position, Dr. White served as a Scientific Advisor with the American Petroleum Institute where she managed toxicology research, regulatory response, and product stewardship efforts for the oil and natural gas industry. She has also held positions as an Environmental Manager for Boar's Head Provisions Co., Inc. and as an Environmental Scientist for Resource Management Concepts, where she managed environmental compliance and sustainability efforts. Dr. White possesses B.S. and M.S. degrees in biology and a Ph. D in Environmental Toxicology. August 24, 2016 Wendy Cleland-Hamnett Director, Office of Pollution Prevention and Toxics Environmental Protection Agency 1200 Pennsylvania Ave. NW Washington, DC 20460-0001 Sent electronically to www.regulations.gov docket # EPA-HQ-OPPT-2016-0400 Re: ACC Comments to Inform EPA"s Rulemaking on the Conduct of Risk Evaluations under the Lautenberg Chemical Safety Act Dear Ms. Cleland-Hamnett: The American Chemistry Council (ACC)1 appreciates the opportunity to provide input to the Office of Pollution Prevention and Toxics to inform the Agency"s development of a risk evaluation rulemaking under the Frank R. Lautenberg Chemical Safety for the 21st Century Act (LCSA). ACC has a long-standing commitment to a robust, science-based approach to evaluation of human and environmental risk. ACC is committed to the effective implementation of the LCSA and supports a workable, rigorous process that allows for timely, high quality reviews. Given the strong emphasis on a risk-based approach in the LCSA, the Section 6(b)(4) rulemaking is particularly important because it will guide the conduct of future risk evaluations that will then inform risk management activities. ACC is committed to being a constructive stakeholder throughout the implementation of LCSA. We will continue to draw from the breadth and depth of our member companies" expertise to ensure that our recommendations are not only science-based, but also allow for the efficient and effective implementation of the LCSA. In doing so, ACC will continue to consider the high quality science standards in the LCSA as well as the timeframes and deadlines imposed therein. The enclosed recommendations were developed with these important considerations in mind. If EPA has any questions, please contact me at nancy_beck@americanchemistry.com or 202-249-6417. Sincerely, Nancy B. Beck, PhD, DABT Senior Director, Regulatory and Technical Affairs Cc: Jim Jones, OCSPP Assistant Administrator Louise Wise, Deputy Assistant Administrator Jeffery Morris, Deputy Director for Programs, OPPT Tala Henry, Director, Risk Assessment Division, OPPT 1 The American Chemistry Council (ACC) represents the leading companies engaged in the business of chemistry. More information about ACC is presented in the body of our comments. 1|Page American Chemistry Council Initial Input to U.S. Environmental Protection Agency In Regard to the Risk Evaluation Rule under the Lautenberg Chemical Safety Act Table of Contents I. Introduction and Executive Summary 4 II. The Risk Evaluation Rulemaking Must Include both Procedural And Substantive Elements to Effect the Purposes of the Statute 5 III. The Proposed Rule Should Include a Tiered Approach to Risk Evaluation 6 IV. The Rule Should Clarify the Process for Preparation and Contents of the Scope 8 V. The Proposed Rule Should Include a Detailed Description of Substantive VI. Elements of Risk Evaluation 8 Proposed Rule Should Ensure Consistency with Section 6(b)(4)(F) 9 a. Integration and Assessment of Information Relevant to Risks and Information on Potentially Exposed and Susceptible Populations 10 i. Conditions of Use That are Relevant 10 ii. Potentially Exposed or Susceptible Subpopulations 10 b. Aggregate and Sentinel Exposures 12 i. Aggregate Exposures 12 ii. Sentinel Exposures 12 c. Exposure Assessment 13 d. Weight of the Evidence 14 VII. The Proposed Rule Should Incorporate Section 26(h) Scientific Standards 14 a. Fit-for-Purpose Approach 15 b. Consideration of Relevant Information 16 i. Improving Hazard Assessment 16 ii. Improving Dose Response Assessment 17 iii. Reliance on Guidance 17 c. Importance of High Quality Risk Characterization 18 d. Clearly Addressing Variability and Uncertainty 18 e. Ensuring Appropriate Peer Review and Forming a Science Advisory Committee on Chemicals 19 2|Page VIII. The Proposed Rule Should Implement a Weight of the Scientific Evidence (WoE) Approach 20 a. Systematic Review is Required 20 i. Development of a Protocol 21 ii. Search Strategy 21 iii. Transparency 21 b. A Systematic Review is Not Automatically a WoE Assessment 21 c. WoE and Systematic Review for Screening Level Risk Evaluations 22 d. WoE and Systematic Review for Refined Risk Evaluations 22 e. Strength of Evidence is Not the Same as WoE 23 IX. EPA Should Make Information Available Consistent with Section 26(j) 24 X. EPA Should Use Reasonably Available Information and CBI Consistent with Section 26(k) 24 XI. EPA Should Utilize Fit-for-Purpose Exposure Evaluation Tools 25 XII. The Requirements of Sections 6 and 26 Apply to Environmental Risk Evaluations 26 a. Advancing Models for Environmental Risks 27 b. Improving Data Sourcing, Generation, and Evaluation 27 c. Persistent, Bioaccumulative and Toxic (PBT) Substances 28 EPA Should Leverage International and Inter-Agency Cooperation 28 XIII. XIV. Incorporating High Throughput Tools and Alternative Methods 29 XV. 30 Stakeholders and EPA Must Be Held to the Same High Standard APPENDIX A: ACC"s Principles for Improving Chemical Hazard and Risk Assessment 31 APPENDIX B: Improving Hazard Assessment 32 APPENDIX C: Improving Does Response Assessment 33 APPRNDIX D: Improving Risk Characterization 35 APPENDIX E: Ensuring Robust Peer Review 37 APPENDIX F: Exposure Modeling Tools 39 APPENDIX G: Additional Information on the ECETOC TRA 41 3|Page I. Introduction and Executive Summary The American Chemistry Council (ACC)2 is pleased to provide the U.S. Environmental Protection Agency (EPA) this initial input on the Lautenberg Chemical Safety Act"s (LCSA) requirement for the Agency to establish, by rule, the process for conducting risk evaluations. ACC appreciates EPA"s early efforts to obtain input from stakeholders at its August 9, 2016, public meeting. We also appreciate EPA"s solicitation of written comments to be entered into the docket, well in advance of publication of the proposed rule. Our comments both clarify, as well as supplement and expand upon, the oral comments we presented at the August 9 meeting. ACC strongly supported Congress" efforts to update and reform the Toxic Substances Control Act (TSCA). We believe that high quality risk evaluation, using best available science and weight of the evidence (WoE), is at the very heart of the LCSA. Effective and efficient risk evaluations will help deliver the results intended by Congress. Section 6(b)(4)(B) of the statute requires EPA to establish, by rule, "a process to conduct risk evaluations." This certainly should include a description of the sequence of events, timelines, opportunities for public comments and peer review. Both Sections 6 and 26 of the LCSA outline various substantive elements that apply to and inform risk evaluation. A risk evaluation must: ? ? ? ? Be conducted in a manner designed to determine "whether a chemical substance presents an unreasonable risk of injury to health or the environment;" as set out in Section 6(b)(4)(A); Identify whether there exists "an unreasonable risk to a potentially exposed or susceptible subpopulation." EPA must identify potentially exposed or susceptible subpopulations relevant to the risk evaluation under conditions of use; Address the specific elements set out in Section 6(b)(4)(F); and Comply with the specific requirements of Section 26, including the best available science, weight of the evidence, and transparency requirements. Because these elements are at the core of the risk evaluation process, and affect risk management measures, they are substantive and should be described in adequate detail in the regulation. In general, where risk evaluation elements are now required by statute, EPA should apply them uniformly and universally reflecting them in the body of the regulation. The recommendations provided by ACC in these comments address screening and refined risk evaluations and are meant to apply to both human health and environmental risks. Specific tools, testing methods, databases, and the like may develop over time, or course, and can be updated as necessary in policies, 2 The American Chemistry Council (ACC) represents the leading companies engaged in the business of chemistry. ACC members apply the science of chemistry to make innovative products and services that make people's lives better, healthier and safer. ACC is committed to improved environmental, health and safety performance through Responsible Care (R), common sense advocacy designed to address major public policy issues, and health and environmental research and product testing. The business of chemistry is a $797 billion enterprise and a key element of the nation's economy. It is the nation"s largest exporter, accounting for fourteen percent of all U.S. exports. Chemistry companies are among the largest investors in research and development. Safety and security have always been primary concerns of ACC members, and they have intensified their efforts, working closely with government agencies to improve security and to defend against any threat to the nation"s critical infrastructure. 4|Page procedures and guidance. Our comments strive to make these differentiations and explain where particular elements of risk evaluation should be included in the rule proper. Specifically, our recommendations suggest definitions, and procedural steps and elements that will allow EPA to ensure that risk evaluations are consistent with the statutory requirements for EPA to use the best available science and WoE approaches. The recommendations also include definitions and procedural steps are not expected to change over time. ACC has referenced each of our suggestions to an existing EPA guidance, a National Academies (NAS) report, or another authoritative body or peer reviewed report. For instance, the recommendations in EPA"s 2000 Risk Characterization Handbook still represent best practices today. Adding adequate definitions and explanation to the rule is particularly important to achieving incorporation of statutory requirements. We also note that in addition to Section 6, Sections 26(h), 26(i), 26(j), and 26(k) of the LCSA each present legal requirements that are applicable to the risk evaluation. EPA will now need to provide a level of transparency regarding not only the inputs, but also the methods of the analysis, including clear descriptions of uncertainties and variability. EPA should leverage information from other jurisdictions where data and information is applicable and of sufficient quality to meet the science standards in the LCSA. Incorporating these elements into the rulemaking creates a better platform for clear and consistent articulation of the Agency"s understanding of statutory requirements, and will better support consistent and uniform application of the elements of risk evaluation. It is critically important that EPA engage the public as EPA plans, scopes, and conducts risk evaluations. Industry scientists often have unique insight and experience with their companies" chemistries and collectively have a large body of knowledge of risk assessment processes globally, including an understanding of potential human health and environmental impacts. ACC encourages EPA to leverage this knowledge and engage early (well before draft risk evaluations are released) and frequently with industry throughout the risk evaluation process. II. The Risk Evaluation Rulemaking Must Include both Procedural and Substantive Elements to Effect the Purposes of the Statute Congress included a specific mandate to EPA to establish a risk evaluation rulemaking. There is little question that the rule must describe the process by which risk evaluations will be conducted.3 However, to effect the purposes of the statute, the process described in the rule cannot merely set out timelines or the sequence of the risk evaluation. It must include a clear articulation of the substantive elements of risk evaluation, and more particularly, it must explain how it will apply the principles set out in Section 6(b)(4)(F), Section 26, and other parts of the statute. If Congress had intended the scientific standard of "best available science" or "weight of the scientific evidence" to be incorporated into guidance alone, it would have included them only in Section 26(l) on "policies, procedures and guidance." 3 "[T]he Administrator shall establish, by rule, a process to conduct risk evaluations in accordance with subparagraph (A)..." Section 6(b)(4)(A). 5|Page The very purpose of the risk evaluation is to develop the evidentiary and scientific basis to enable EPA to complete the risk determination required by statute. That risk determination has substantive impact - it significantly affects conduct, activity or a substantive interest that is the subject of agency regulation. The determination following risk evaluation is a necessary prerequisite for a chemical to proceed to risk management, if warranted. The rule should thus include a clear description of how EPA will undertake risk evaluations in order to meet the new statutory requirements of the LCSA. This includes a description of the scoping process and requirements for a published scope as well as the elements of the risk evaluation itself and the mechanism for gauging adequacy as measured against statutory criteria. III. The Proposed Rule Should Include a Tiered Approach to Risk Evaluation We believe the statute contemplates a tiered approach to risk evaluation and recommend that EPA include a tiered approach in the rule. Under the LCSA, EPA must initiate the risk evaluation "upon designating" a chemical as a high-priority substance. The scope, however, is not required to be published "upon initiation" -- EPA has up to six months following the initiation of the risk evaluation to prepare and publish the scope. Congress intended this six month period to be used for a scoping exercise, where EPA identifies "the hazards, exposures, conditions of use, and the potentially exposed or susceptible subpopulations the Administrator expects to consider in the risk evaluation." This six month period is a "step" between the high priority designation and the publication of the scope. In order for EPA to conduct risk evaluations consistent with the quality required by the LCSA and within the timeframes required, EPA should conduct a screening level evaluation during the scoping phase. During the scoping phase of risk evaluations, tools exist to allow EPA to conduct quantitative screening level analyses of multiple exposure scenarios, as appropriate for consumers, sensitive subpopulations, and the environment. This will allow EPA to have a more tailored focus on those populations and exposures of greatest concern during a refined risk evaluation process. Figure 1 below depicts ACC"s recommended approach. 6|Page Figure 1. A Two-Step Process for Conducting Risk Evaluations Note: This is a simplified version of the process. A tiered approach, where EPA uses the scoping step (step 1) to conduct a quantitative screening level analysis, will allow EPA to focus its limited resources on more robust refined risk evaluations for only those conditions of use where unreasonable risks cannot be ruled out. Screening-level assessments require less data and information, and are typically deterministic and based on conservative, health protective assumptions and methods. When a screening assessment indicates low risk for a particular condition of use, the Agency should have a high degree of confidence that the potential risks are much lower than the calculation and, therefore, the actual risks are lower and/or perhaps non-existent. However, when a screening-level risk assessment indicates a potential concern for an adverse effect, this does not mean that the actual risks are significant and warrant action. Rather, it indicates the Agency should take a second step in the risk evaluation process to refine the evaluation to more accurately quantify potential risks. The refined risk evaluation (step 2) will require realistic and representative data, higher tier modeling approaches, including probabilistic exposure modeling, and a more comprehensive consideration of human relevance and dose-response relationships. In a refined evaluation, EPA should also consider targeted exposure studies, as well as biomonitoring and environmental monitoring data, to the extent that this information is available and relevant. This approach is consistent with EPA"s 2014 Framework for Human Health Risk Assessment to Inform Decision Making (HHRA Framework)4, which also emphasizes the importance of a fit-for-purpose approach to risk evaluation. This approach is also consistent with EPA"s exposure assessment guidelines and practices.5 The concept of a tiered approach and a fit-for-purpose evaluation are woven throughout ACC"s recommendations. 4 5 See https://www.epa.gov/sites/production/files/2014-12/documents/hhra-framework-final-2014.pdf. See: https://www.epa.gov/expobox/exposure-assessment-tools-tiers-and-types-screening-level-and-refined. 7|Page The tiered approach ACC recommends is consistent with the approach EPA took in the problem formulation and initial assessment document for tetrabromobisphenol A (TBBPA).6 In that document, EPA conducted an initial screening level evaluation to support its conceptual model and analysis plan. EPA appropriately used high-end exposure values coupled with the lowest toxicity values to evaluate uses and exposure pathways of potential concern. While EPA did not share the relevant risk evaluation calculations in its public document, the general approach is consistent with that of a screening level risk evaluation. ACC encourages EPA to continue with this approach and to transparently and clearly present quantitative screening level analyses for the conditions of use and exposure scenarios that are part of the conceptual model EPA develops as part of the scoping phase. IV. The Rule Should Clarify the Process for Preparation and Contents of the Scope As noted above, Congress allowed a six month period for preparation of the scope of the risk evaluation, contemplating that time and effort would be needed to move from prioritization to a published scope. The six month period is to enable EPA to identify "the hazards, exposures, conditions of use, and the potentially exposed or susceptible subpopulations the Administrator expects to consider in the risk evaluation." Two things are evident from this language and the time frame afforded: 1) EPA should use this period to evaluate and decide which, if any, potentially exposed or susceptible subpopulations should be included in the risk evaluation (in other words, it need not include all such subpopulations, regardless of size, impact, or relevance); and 2) tEPA has flexibility to actually conduct a full risk evaluation of some or all the potential scenarios set out in the scope. In short, EPA need not include every conceivable condition of use in a risk evaluation. This view is further buttressed by the definition of "conditions of use" in Section 3 of the LCSA, which points to the need for EPA to determine the relevant conditions of use: "the circumstances, as determined by the Administrator, under which a chemical substance is intended, known, or reasonably foreseen to be manufactured, processed, distributed in commerce, used, or disposed of." (Emphasis added). V. The Proposed Rule Should Include a Detailed Description of Substantive Elements of Risk Evaluation The term, "risk evaluation" is not expressly defined in the LCSA. While the term "risk assessment" has been widely used in EPA programs and operationally has clear meaning derived from years of guidance, policies and practices, that term was not used in the statute. Therefore even though it may be reasonable to assume "risk evaluation" may fully equate with the term "risk assessment," given the context of its use (integrating hazard with exposure) in the LCSA, EPA is encouraged to explicitly define and operationalize this term as part of its rulemaking. The term will not have clear meaning until an interpretation is assigned by EPA. We believe the essential elements of a Section 6 and 26 risk evaluation must be articulated in a clear regulatory definition as we discuss below. 6 EPA, Problem Formulation and Initial Assessment Tetrabromobisphenol A and Related Chemicals Cluster Flame Retardants, 2015, available at: https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/tsca-work-plan-chemical-problemformulation-and-2. 8|Page Section 6(b)(4)(B) of the statute requires EPA to establish, by rule, "a process to conduct risk evaluations." This process is itself required to meet a number of substantive elements described in the LCSA; a risk evaluation must: o o o o Be conducted in a manner designed to help the agency determine "whether a chemical substance presents an unreasonable risk of injury to health or the environment;" as set out in Section 6(b)(4)(A). Include consideration of "an unreasonable risk to a potentially exposed or susceptible subpopulation." EPA must identify relevant potentially exposed or susceptible subpopulations relevant to the risk evaluation under conditions of use; Address the specific elements set out in Section 6(b)(4)(F); and Comply with the specific requirements of Section 26, including the best available science, weight of the evidence, and transparency requirements. The very purpose of the risk evaluation is to develop the evidentiary and scientific basis to enable EPA to complete the risk determination required by statute. That risk determination has substantive impact - it significantly affects conduct, activity or a substantive interest that is the subject of agency regulation. The basis for the risk determination thus should be adequately described in the rule itself to offer sufficient notice to the regulated community. This is particularly important for decisions that inform safety and safety determinations.7 Likewise, decisions that have broad reaching impact should be supported in regulations, not merely through guidance or agency policy.8 While EPA cannot substitute policy or guidance for a regulatory description of what will constitute a complete and robust risk evaluation, we believe the necessary elements can be developed in this rulemaking in a timely manner. VI. The Proposed Rule Should Ensure Consistency with Section 6(b)(4)(F) As discussed below, Section 6(b)(4)(F) of the LCSA describes five requirements for risk evaluations that shall be considered by the Administrator and must be incorporated into the risk evaluation rulemaking. 7 See, e.g., MST Express v. U.S. Department of Transportation, 108 F.3d 401 (D.C. Cir. 1997). DOT was directed under the Motor Carrier Safety Act (MCSA) to "prescribe regulations establishing a procedure to decide on the safety fitness of owners and operators of commercial motor vehicles." [Emphasis added]. The MCSA stated that implementing regulations would include "a means of deciding whether the owners, operators, and persons meet the safety fitness requirements." DOT promulgated regulations that set out a process for decision making but used guidance to articulate the tests by which the agency would determine whether vehicles met the safety fitness requirements. The court rejected DOT"s reliance on guidance because it "failed to carry out its statutory obligation to establish by regulation a means of determining whether a carrier has complied with the safety fitness requirements." 8 As a general matter, "...it seems to be established that ,,regulations," ,,substantive rules" or ,,legislative rules" are those which create law, usually implementary to an existing law." Gibson Wine Co. v. Snyder, 194 F.2d 329, 331 (D.C. Cir. 1952), cited by Brown Express, Inc. v. U.S., 607 F.2d 695, 700 (5th Cir. 1979). A "rule" is defined under Section 2 of the Administrative Procedure Act, in relevant part, as: "the whole or part of an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy or describing the organization, procedure, or practice requirements of an agency." 5 U.S.C. ? 551(4). 9|Page a. Integration and Assessment of Information Relevant to Risks and Information on Potentially Exposed and Susceptible Populations Section 6(b)(4)(F)(i) requires risk evaluations to integrate and assess available information on hazards and exposures for the conditions of use of the chemical substance. The statute does not, however, explain how information will be integrated or how it will be assessed. The process of how and when information will be integrated and assessed should be described in the proposed rule. There are two key considerations in section 6(b)(4)(F)(i).9 First, EPA must integrate and assess information for conditions of use that are relevant to risks to human health and the environment, and second, EPA must provide information on potentially exposed or susceptible subpopulations identified as relevant. i. Conditions of Use That are Relevant As discussed above, had Congress intended EPA to necessarily address all "conditions of use" in an evaluation, that term (and EPA"s ability to define the conditions of use) would have been surplusage. During the scoping step of the risk evaluation process, a screening level risk evaluation will allow EPA to focus on those conditions of use that may pose a potential risk to human health and/or the environment under relevant exposure conditions. EPA should consider only conditions of use which are "intended, known, or reasonably foreseen to be manufactured, processed, distributed in commerce, used, or disposed of." EPA should not include exposure scenarios that are in clear violation of OSHA workplace limits or EPA regulatory requirements, exposures that are not consistent with labeling requirements for safe use, or exposures that are inconsistent with intended uses of consumer products. In addition, EPA should not include exposure scenarios regulated under other federal laws. ii. Potentially Exposed or Susceptible Subpopulations Section 6(b)(4)(F)(i) requires EPA to integrate and assess information on potentially exposed or susceptible subpopulations identified by EPA as relevant to the risk evaluation. In the proposed rule, EPA should describe the process it will use to identify subpopulations and how it will make a determination that the subpopulation is relevant to the risk evaluation. The statute does offer a definition of "potentially exposed or susceptible subpopulation" in Section 3(12) that clarifies that the subpopulation must have either "greater susceptibility" or "greater exposure" to a chemical substance or mixture, and due to this "may be at greater risk" than the general population.10 While this is helpful, it is not sufficient to inform the risk evaluation process. 9 Section 6(b)(4)(F)(i) states that EPA must "integrate and assess available information on hazards and exposures for the conditions of use of the chemical substance, including information that is relevant to specific risks of injury to health or the environment and information on potentially exposed or susceptible subpopulations identified as relevant by the Administrator." 10 Section 3(12) states "The term ,,potentially exposed or susceptible subpopulation" means a group of individuals within the general population identified by the Administrator who, due to either greater susceptibility or greater exposure, may be at greater risk than the general population of adverse health effects from exposure to a chemical substance or mixture, such as infants, children, pregnant women, workers, or the elderly." 10 | P a g e The term "potentially exposed" would benefit from EPA"s clarification in the body of the rule. It modifies "subpopulation," and we do not believe Congress intended "potentially exposed subpopulation" to mean any subpopulation, without boundaries, for which potential exposure could be conceived. A subpopulation presumably does not mean any two people anywhere who might be exposed to as little as one molecule of a chemical substance for a millisecond pursuant to a non-authorized use; or that a predator animal will spend its entire life receiving exposures from a single soil boring. An expansive view of potential exposure would make risk evaluations difficult to scope, would defeat the purposes of disciplined and methodical risk evaluation, and would impede their timely completion. Such an approach also runs counter to Congressional intent that the Agency focus on unreasonable risk, not every conceivable risk to every conceivable population. Thus, a risk evaluation should exclude potentially exposed subpopulations where the potential exposure is such that negligible risks are implicated. It would be helpful to define the term in the in the proposed rulemaking to ensure alignment with the LCSA intent. We also note that the term "potentially exposed" would benefit from regulatory alignment with the "greater exposure" provision. The first term concerns itself with potential exposure and the second with actual exposure (and for that matter, actual exposure that exceeds the actual exposure of the general population). The proposed rule should offer this clarification. To integrate Section 26 science requirements, an explanation can be added in the proposed rule to explain what constitutes best available exposure science for purposes of reaching an agency decision as to what "potentially exposed" means. For example, if a statistical probability of exposure is calculated or exposure modeling used, or biomonitoring data is used in a subpopulation as a proxy for exposures, the decision to use a particular approach should be outlined in the risk assessment and accompanied with an explanation of why the approach satisfies the best available science requirement. The term "greater susceptibility" would also benefit from a regulatory definition. The definition of "potentially exposed subpopulation" offered in the statute does make clear that "greater susceptibility" is a different concept than "greater exposure," but it would also be helpful for EPA to explain that "greater susceptibility" means groups that respond differently (presumably, with greater potential for adverse health effects, or due to greater sensitivity) to the same chemical exposure. It would also be helpful for EPA to explain that its determination that a group has "greater susceptibility" cannot be based on assumption, speculation, or anecdotal evidence, but must be founded on high quality science consistent with the best available science, weight of the evidence, and other requirements of Section 26. We also encourage EPA to consider defining the term "group" and assign a reasonable floor or range to group size based on accepted criteria. Micropopulations of groups as small as two or three individuals, while conceivable in the abstract, would generally not exist or be identifiable in a risk evaluation exercise. It would be helpful for EPA to clarify in the rule that the term "group" should be read together with "subpopulation" to identify categories of individuals that are statistically large enough to justify identification and support the throughput objectives of the statute for risk evaluations. Examples of subpopulations identified in epidemiological and exposure science may be particularly helpful here. 11 | P a g e b. Aggregate and Sentinel Exposures While the plain language of the statute makes clear that aggregate or sentinel exposures need not be considered at all in the risk evaluation, EPA has discretion to do so.11 If they are considered, Section 6(b)(4)(F)(ii) requires EPA to describe the basis for that consideration. We believe EPA should include definitions of both aggregate and sentinel exposures in the proposed regulation so the regulated community will understand how EPA intends to apply the terms, and to ensure consistency and regularity in application. i. Aggregate Exposures Consistent with EPA"s exposure toolbox,12 EPA should continue to define aggregate exposures as the combined exposure for one substance over multiple exposure pathways from multiple different sources. As noted above, while EPA is not required to conduct aggregate exposure assessments, aggregate exposure assessment may be appropriate when available data are of sufficient quality, reliable, and representative of intended, known or reasonably foreseen exposures. ii. Sentinel Exposures A sentinel exposure should be thought of as the exposure that is judged to cause the plausible upper-bound individual human exposure to a substance of interest within a broad category. For example, a broad category would be air care products, which contains subcategories of air care instant action (aerosol sprays) and air care continuous action (solids and liquids) products. If the estimated exposure for the sentinel product in a broad category results in an acceptable risk assessment outcome when compared with the appropriate hazard reference value, then there is no need to continue estimating the exposure for the different subcategories. Evaluating sentinel exposures first, during a screening level evaluation in the scoping phase, will allow EPA to focus on exposures of the greatest relevance and importance, within a particular category of exposures. This is not intended to mean, for instance, that high-end occupational exposure is sentinel for consumer exposure. Rather, it can be used in an overall Product Category context - where a sub-category with highest exposure can represent or be the sentinel for other subcategories. If the sentinel sub-category is adequately protected, then the rest are seen as acceptable. This approach is consistent with the approaches and guidance of ECHA for REACH chemical evaluations13 and the European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC).14 11 Section 6(b)(4)(F)(ii) provides that EPA must "describe whether aggregate or sentinel exposures to a chemical substance under the conditions of use were considered, and the basis for that consideration." 12 See: https://www.epa.gov/expobox/exposure-assessment-tools-tiers-and-types-aggregate-and-cumulative. 13 See ECHA Guidance on information requirements and chemical safety assessment Chapter R.15: Consumer exposure estimation, November 2009 (version 2 Rev.:0.0), available at: https://echa.europa.eu/documents/10162/13632/r15_update_version_2_rev00_en.pdf. 14 See Addendum to ECETOC Targeted Risk Assessment Report No. 93, 2009, available at: http://www.ecetoc.org/wpcontent/uploads/2014/08/ECETOC-TR-107-Addendum-to-ECETOC-TRA-report-93.pdf. 12 | P a g e c. Exposure Assessment Congress" intent to ensure a high quality exposure assessment is made clear in Section 6(b)(4)(F)(iv) and EPA must ensure that these quality characteristics are incorporated into the rulemaking.15 Understanding duration, intensity, and frequency are key elements of ensuring that the exposure assessment is robust and goes beyond the screening level. However, as covered in multiple authoritative reports and guidance documents, when considering exposure, either to humans or the environment, there are many other aspects that EPA must consider to be consistent with the high quality requirements of both Section 6 and 26 of the LCSA. For instance, consistent with EPA exposure guidance, exposure should also consider physicochemical properties, distribution, and fate in the environment, including products of metabolism and chemical break down products.16 Sufficient, reliable data should be used over default assumptions and if models are used to estimate exposure, their strengths and limitations must be clearly described and sufficient information must be made available to enable others to replicate and verify the modeling.17 EPA must take extra care to provide clear rationales for the use of any default assumptions.18 To ensure consistency with Section 26 requirements, and best practices, EPA must clearly define uncertainties and conduct a sensitivity analysis to evaluate the impacts of individual parameters on the exposure conclusions.19 In addition, to be consistent with a WoE approach, all sources of exposure information must be described and evaluated for quality and reliability, to ensure reliance on the highest quality information. As mentioned previously, for refined risk evaluations, consistency with the best available science means that EPA should strive to use probabilistic approaches and the presentation of results should ensure that distributions of exposure, including central estimate and reasonable maximum exposure (RME) estimates are clear.20 Pharmacokinetic information, biomonitoring information, and environmental monitoring information should be integrated into the assessment when it is of sufficient quality and available.21 Finally, if only minimal information is available to assess exposure, EPA should identify any additional information needs.22 Consistent with the Section 26 requirement to use "best available science," when evaluating exposure information, we encourage EPA to use probabilistic approaches that rely on high quality data. A probabilistic approach will prevent over-reliance on data points that represent the tails of distributions and may not be stable or reproducible. In the presentation on August 9, 2016, EPA noted that for occupational exposures, high-end estimates of exposure would include the 95th or 15 Section 6(b)(4)(F)(iv) states that the Administrator shall "take into account, where relevant, the likely duration, intensity, frequency, and number of exposures under the conditions of use of the chemical substance." 16 EPA Guidelines for Exposure Assessment, 1992 ("EPA Exposure Guidelines"), available at: https://www.epa.gov/sites/production/files/2014-11/documents/guidelines_exp_assessment.pdf. 17 Ibid.; Fenner-Crisp PA, Dellarco VL., Key Elements for Judging the Quality of a Risk Assessment, Environ Health Perspect. 2016 Feb 5, available at: http://ehp.niehs.nih.gov/wp-content/uploads/124/8/ehp.1510483.alt.pdf. 18 Ibid. 19 Ibid. 20 National Research Council, National Academy of Sciences, Models in Environmental Regulatory Decision Making, 2007, available at http://www.nap.edu/catalog/11972/models-in-environmental-regulatory-decision-making; Fenner-Crisp and Dellarco, 2016. 21 EPA Exposure Guidelines. 22 Ibid.; Fenner-Crisp and Dellarco, 2016. 13 | P a g e 99th percentiles.23 We note that, for occupational exposures under REACH, ECHA finds the 75th or 90th percentiles to be appropriate, particularly focusing on the 75th percentile for two situations: when measured data may represent worst case exposure activities rather than typical (air monitoring is usually performed on the tasks with highest potential for exposure) or when measured data may be high quality and represent narrow exposure conditions with consistent operational conditions and risk management measures.24 When conducting refined risk evaluations, ACC encourages EPA to make these determinations using situation specific information to produce estimates that are realistic and reliable. We also encourage EPA to continue to improve the transparency and clarity of exposure assessments, ensuring that all default assumptions and uncertainties are clearly articulated. d. Weight of the Evidence Section 6(b)(4)(F)(v) requires a description of the WoE evaluation for both hazard and exposure and Section 26(i) requires that EPA make decisions under sections 4, 5, and 6 "based on the weight of the scientific evidence."25 Section VIII below will discuss this requirement in further detail. VII. The Proposed Rule Should Incorporate Section 26(h) Scientific Standards Risk evaluations - and the risk evaluation process - must satisfy the specific requirements of Section 6, but also must satisfy the requirements of Section 26. Section 26 must be read in tandem with Section 6 requirements to effect the purpose and intent of the LCSA and must be incorporated into the rulemaking. Risk evaluations and the risk evaluation process must comply with the best available science provision in Section 26(h) of the LCSA, as well as the WoE provision at 26(i) and the transparency provision at 26(j). While "best available science" is not explicitly defined in the LCSA, the intent is made clear as subsections of section 26(h) and section 26(i) are fully consistent with the descriptions provided in the Senate Report.26 23 See slide 36 available at: https://www.epa.gov/sites/production/files/2016-08/documents/risk_evaluation_9_august_2016.pdf. See ECHA Guidance on Information Requirements and Chemical Safety Assessment Part R.14: Occupational exposure assessment Draft (Public) Version 3.0 November 2015, at page 23, available at: https://echa.europa.eu/documents/10162/13564/r14_draft_for_peg_en.pdf. This is also consistent with the most recent final consultation update available at: https://echa.europa.eu/documents/10162/13564/r_14_caracal_en.pdf/18442141-4b1a-41cbb2eb-c619aae9fcb5. 25 Section 6(b)(4)(F)(v) states that the Administrator shall "describe the weight of the scientific evidence for the identified hazard and exposure." Section 26(i) states "The Administrator shall make decisions under sections 4, 5, and 6 based on the weight of the scientific evidence." 26 Senate Report 114-67 states, at page 8-9, "It is the Committee"s intent that decision-making under TSCA be ,,consistent with the best available science." The science EPA relies on to make safety determinations should describe and document any assumptions and methods used, and should address variability, uncertainty, the degree of independent verification and peer review. The section also requires that decisions be based on the weight of the scientific evidence, by which the Committee intends that EPA consider all information in a systematic and integrative framework to consider the relevance of different information. The Committee believes there is significant value, where available and appropriate, in EPA"s use of peer reviewed information, standardized test design and methods, consistent data evaluation procedures and good laboratory practices to ensure transparent, understandable, and reproducible chemical reviews." Available at: https://www.congress.gov/114/crpt/srpt67/CRPT-114srpt67.pdf. 24 14 | P a g e OMB Information Quality Guidelines (IQ Guidelines) set forth special considerations that generally apply to risk assessments. The IQ Guidelines state: "With regard to analysis of risks to human health, safety, and the environment maintained or disseminated by the agencies, agencies shall either adopt or adapt the quality principles applied by Congress to risk information and disseminated pursuant to the Safe Drinking Water Act Amendments of 1996 (SDWA)." The SDWA principles require, to the extent that agency action is based on science, that the agency use "the best available, peer-reviewed science and supporting studies conducted in accordance with sound and objective scientific practices."27 EPA adopted the SDWA principles in its Information Quality Guidelines, and we encourage regulatory adoption here. Consistent with Section 26 of the LCSA, it is imperative that the definition incorporated in the rulemaking also explicitly address the needs for high quality, relevancy, transparency, and WoE evaluations. Beyond including a definition of best available science in the rulemaking itself, EPA should include in its proposed rule the elements that will be considered in a best available science review. We encourage EPA to review Fenner-Crisp (2016), Key Elements for Judging the Quality of a Risk Assessment, 28 the ACC Principles for Improving Chemical Hazard and Risk Assessments (see Appendix A),29 and other principles and elements described in these comments, to illustrate best available science for inclusion in the proposed rulemaking. These principles and practices are not expected to change over time. As such, EPA should have no concerns incorporating them into the risk evaluation rulemaking. a. Fit-for-Purpose Approach Section 26(h)(1) is intended to ensure that the information, approaches, methods and models used are appropriate for the intended use of the information.30 Consistent with the two-step tiered framework ACC proposes (discussed in Section III, above) where a screening level evaluation is conducted in the scoping phase, and a more refined risk evaluation is conducted to directly inform risk management activities, when considering the best available science, EPA must take its "fit-forpurpose" use into account. EPA needs to exercise care when including or excluding information. There may be a bias in favor of including information and data simply because it has been peer reviewed or is performed under GLP. There may also be a bias against information and data which is not. Neither bias is correct. EPA will need to consider other factors that may also be important and contribute to identifying best-available and high quality information for the specific evaluation being conducted. A default-based, conservative model may be appropriate for a screening evaluation. However, the best available science for a refined risk evaluation requires high quality, reliable data be used over defaults, when available, and that any default assumptions that are used must be appropriate for the decision being made. This approach is well outlined in EPA"s HHRA Framework.31 The 2016 Environmental Health Perspectives (EHP) publication by Drs. Fenner-Crisp and Dellarco refers back to recommendations made as early as 1997 where the Presidential/Congressional Commission 27 42 U.S.C. ? 300g-1(b)(3)(A,B). Fenner-Crisp PA, Dellarco VL., Key Elements for Judging the Quality of a Risk Assessment, Environ Health Perspect. 2016 Feb 5, available at: http://ehp.niehs.nih.gov/wp-content/uploads/124/8/ehp.1510483.alt.pdf 29 ACC Principles for Improving Chemical Hazard and Risk Assessments available at: https://www.americanchemistry.com/WorkArea/DownloadAsset.aspx?id=8309. 30 Section 26(h)(1) states that the Administrator shall consider "the extent to which the scientific information, technical procedures, measures, methods, protocols, methodologies, or models employed to generate the information are reasonable for and consistent with the intended use of the information." 31 See https://www.epa.gov/sites/production/files/2014-12/documents/hhra-framework-final-2014.pdf. 28 15 | P a g e on Risk Assessment and Risk Management discussed the need for the complexity and depth of an assessment to be commensurate with the decision being made.32 ACC encourages EPA to follow the overarching process that is described in these documents. b. Consideration of Relevant Information Section 26(h)(2) stresses the importance of the consideration of relevant information.33 That is, the scientific information must be relevant to human or environmental health. This is particularly important for the refined risk evaluations that will be used to inform risk management controls for particular uses. These controls could be as rigorous as a ban, or a costly labeling requirement, thus the highest quality data and methodologies must be used. Two areas where EPA will need to improve its consideration of relevant information, for refined risk evaluations, are hazard assessment and dose-response. i. Improving Hazard Assessment Sections 26(h) and 26(i) collectively require that the risk evaluation rulemaking ensure improvements in EPA"s current approach to hazard assessment, particularly for those assessments that will be used to inform risk management decisions in Section 6(c). EPA may be tempted to rely on data from existing databases. However, the requirements of the new statute make this very challenging. For instance, it is well known that the EPA IRIS program has been struggling for years to produce high quality assessments. The National Academies (NAS) has commented on this, critically noting "persistent problems" in IRIS assessments.34 While the IRIS program is working to address these systemic problems, the program has, to date, not finalized a single IRIS assessment that is fully consistent with the 2011 recommendations of the NAS. As the IRIS program only releases a few final assessments a year, this means that the majority of information in the IRIS database may be outdated and not representative of the best available science, and the assessment processes used to evaluate, judge, and synthesize information may not be consistent with the best scientific evaluation practices that exist today. All of these concerns have been well articulated by the NAS. While the risk evaluation approach currently used by OPPT does not require the direct adoption of RfDs and RfCs from IRIS, OPPT must be extremely cautious when relying on studies, points-of-departure, or endpoints simply because they were identified in an IRIS assessment. This justification will not suffice under the scientific requirements in the LCSA. Similarly, modeling approaches used by IRIS may not be fully informed by the weight of the evidence, including a review of mode of action information and mechanistic data. While relying on studies and data from IRIS may be appropriate for screening level assessments, OPPT will need to conduct new WoE evaluations for hazard identification in the refined risk evaluation process. Simply selecting the lowest value because it is health protective, regardless of quality and relevance, is not acceptable under the LCSA. 32 Fenner-Crisp and Dellarco, 2016. Section 26(h)(2) states that the Administrator shall consider "the extent to which the information is relevant for the use of the Administrator in making a decision about a chemical substance or mixture." 34 See, for example chapter 7 of the NAS Review of EPA"s IRIS Assessment of Formaldehyde (2011), available at: http://www.nap.edu/catalog/13142/review-of-the-environmental-protection-agencys-draft-iris-assessment-of-formaldehyde. 33 16 | P a g e Fortunately, many authoritative bodies and government documents provide suggestions for the improvement of hazard assessment, including some EPA guidance documents. ACC encourages EPA to incorporate the five elements described in Appendix B "Improving Hazard Assessment," which would provide significant improvements to hazard assessment and help to ensure EPA"s consistency with the LCSA. We recommend that each of these elements be incorporated into the risk evaluation rulemaking. These elements have not changed over time and are not expected to change, thus EPA should not have concerns incorporating them into the risk evaluation rulemaking. ii. Improving Dose Response Assessment As EPA will be providing quantitative estimates of risk to inform the decision of whether risk management is necessary, a high quality dose-response assessment will be an important element of EPA"s refined risk evaluations. Significant guidance exists to provide a framework to ensure that dose-response evaluations are consistent with the LCSA. ACC provides nine recommendations, including references, in Appendix C "Improving Dose Response Assessment," for refined risk evaluations that should be incorporated into EPA"s rulemaking for risk evaluation. These elements of a high quality dose response assessment are not expected to change over time. iii. Reliance on Guidance While significant guidance exists on best practices regarding the process of conducting a risk evaluation, EPA must ensure that any reliance on a guidance document is consistent with the best available science. EPA guidance documents are grounded in principles of public health protection. As discussed by EPA in the 2004 Staff Paper on Principles and Practices, EPA ensures that risk is not likely to be underestimated and default assumptions are used in risk assessments to pursue this goal.35 As such, when EPA refers to guidance documents, particularly older documents, EPA must ensure they are relying on science, not necessarily defaults. If EPA is relying on defaults, EPA must ensure that these default assumptions are still consistent with today"s best available science. For example, in the presentation on August 9, 2016, EPA pointed to the 1991 Guidelines for Developmental Toxicity Risk Assessment. EPA"s current science-policy position is based on an assumption that was articulated in 1991.36 EPA must ensure that the 1991 guidance assumption is still appropriate today; it cannot simply be presumed to be accurate. EPA must now take into account all the new general and chemicalspecific evidence (including pharmacokinetic data), methodological study designs for 35 EPA Staff Paper on Risk Assessment Principles and Practices, 2004, available at: https://nepis.epa.gov/Exe/ZyNET.exe/100045MJ.TXT?ZyActionD=ZyDocument&Client=EPA&Index=2000+Thru+2005&Doc s=&Query=&Time=&EndTime=&SearchMethod=1&TocRestrict=n&Toc=&TocEntry=&QField=&QFieldYear=&QFieldMont h=&QFieldDay=&IntQFieldOp=0&ExtQFieldOp=0&XmlQuery=&File=D%3A%5Czyfiles%5CIndex%20Data%5C00thru05% 5CTxt%5C00000007%5C100045MJ.txt&User=ANONYMOUS&Password=anonymous&SortMethod=h%7C&MaximumDocuments=1&FuzzyDegree=0&ImageQuality=r75g8/r75g8/x150y150g16/i425&Display=hpfr&DefSeekPage=x& SearchBack=ZyActionL&Back=ZyActionS&BackDesc=Results%20page&MaximumPages=1&ZyEntry=1&SeekPage=x&ZyP URL#. 36 EPA Guidelines for Developmental Toxicity Risk Assessment, 1991, at page 38 states, "Second, for developmental toxic effects, a primary assumption is that a single exposure at a critical time in development may produce an adverse developmental effect, i.e., repeated exposure is not a necessary prerequisite for developmental toxicity to be manifested." Available at: https://www.epa.gov/sites/production/files/2014-11/documents/dev_tox.pdf. 17 | P a g e developmental assessment, and study quality criteria that may exist to inform the use of this default approach. c. Importance of a High Quality Risk Characterization Section 26(h)(3) speaks to the importance of clearly documenting the data and approaches used in the risk evaluation, including, but not limited to the methods, assumptions and quality assurance approaches used to generate the information.37 For risk evaluation, this requirement ensures that steps are taken to provide a very clear risk characterization section in each refined risk evaluation. This step must be incorporated into the risk evaluation rule and we strongly encourage EPA to fully implement the recommendations in the 2000 EPA Risk Characterization Handbook.38 ACC has identified key aspects of this handbook that are critical to ensuring that risk evaluations are consistent with the science standards in the LCSA. These recommendations are provided in Appendix D "Improving Risk Characterization." ACC recommends that each of these key aspects, which have been unchanged since 2000, be incorporated into the risk evaluation rulemaking. d. Clearly Addressing Variability and Uncertainty Section 26(h)(4) calls out the importance of clarity and transparency, particularly as it relates to characterizing variability and uncertainty in both the methods and protocols as well as the models and the information.39 EPA must not underestimate the importance of clarity and full transparency when describing this information as it should be a critical consideration when contemplating potential risk management actions. Where feasible, to be consistent with Sections 6 and 26, EPA must characterize uncertainty and variability quantitatively, particularly for the refined risk evaluations. This is fully consistent with a fit-for-purpose approach to risk evaluation. Any limitations in the analysis must be explained clearly, including discussion of the impacts that the limitations may have on the end results. For refined risk evaluations, EPA must ensure that when a quantitative uncertainty analysis is provided, it must be probabilistic and the data, methods, and models used are described sufficiently to allow for independent re-analysis.40 If a quantitative uncertainty analysis is not provided, the omission should be justified and included in the risk evaluation.41 Equally important, EPA must ensure that variability in effects or responses across relevant populations(s) are discussed with significant uncertainties noted.42 In 2013, ACC"s Center for Advancing Risk Assessment Science and Policy (ARASP) hosted an invited participant workshop to explore approaches to improve methods for presenting uncertainty 37 Section 26(h)(3) states that the Administrator shall consider: "the degree of clarity and completeness with which the data, assumptions, methods, quality assurance, and analyses employed to generate the information are documented." 38 EPA Risk Characterization Handbook, 2000, available at: https://www.epa.gov/sites/production/files/201510/documents/osp_risk_characterization_handbook_2000.pdf. 39 Section 26(h)(4) states that the Administrator shall consider "the extent to which the variability and uncertainty in the information, or in the procedures, measures, methods, protocols, methodologies, or models, are evaluated and characterized." 40 EPA Risk Characterization Handbook, 2000; Fenner-Crisp and Dellarco, 2016; OMB/OSTP memorandum "Updated Principles for Risk Analysis," 2007, available at: https://www.whitehouse.gov/sites/default/files/omb/assets/regulatory_matters_pdf/m07-24.pdf. 41 Ibid. 42 NAS Pesticides in the diets of infants and children, 1993, available at: http://www.nap.edu/catalog/2126/pesticides-in-thediets-of-infants-and-children; Fenner-Crisp and Dellarco, 2016. 18 | P a g e and risk information in federal chemical hazard assessment programs. Participants included more than 60 experts in toxicology, risk assessment, risk communication, exposure assessment, and hazard characterization drawn from academia, government (including EPA), and industry, and nongovernmental organizations. ACC encourages EPA to consider the 2016 publication43 that resulted from this workshop to inform improved methods for presenting uncertainty information as it relates to hazard assessment, a critical element of the risk evaluation process. e. Ensuring Appropriate Peer Review and Forming a Science Advisory Committee on Chemicals Section 26(h)(5) recognizes the important role of peer review.44 Consistent with EPA"s own Peer Review Handbook,45 the most robust reviews should be reserved for the refined risk evaluations. ACC expects that EPA will adhere to the principles, guidance and criteria set forth in the OMB Information Quality Bulletin for Peer Review,46 which EPA has incorporated and expanded upon in its Peer Review Handbook. For consistency with this requirement of the LCSA, ACC recommends that the specific peer review process that will be used for refined risk evaluations be clearly described in the risk evaluation rulemaking. This includes elements relating to how reviewers will be selected, transparency of the review process, inclusion of stakeholders and the need for a robust consensus report. As depicted in Figure 1, a robust peer review should be conducted on the draft refined risk evaluation.47 Robust peer review and public comment will serve to improve the quality, credibility, and acceptance of the final risk evaluation. Appendix E "Ensuring Robust Peer Review" provides specific recommendations for incorporation into the rulemaking. ACC also recommends, consistent with Section 26(o) of the LCSA, that EPA expeditiously form a Science Advisory Committee on Chemicals (SACC) and use this group (or a subgroup of this group) to conduct the peer review of refined risk evaluations. EPA"s current Chemical Safety Advisory Committee does not meet the requirements of the LCSA nor is the membership of the committee sufficient to ensure high quality reviews, as ACC has commented in the past.48 43 Beck NB, et al., Approaches for describing and communicating overall uncertainty in toxicity characterizations: U.S. Environmental Protection Agency's Integrated Risk Information System (IRIS) as a case study; Environ Int. 2016, available at: http://www.sciencedirect.com/science/article/pii/S0160412015301367. 44 Section 26(h)(5) states that the Administrator shall consider "the extent of independent verification or peer review of the information or of the procedures, measures, methods, protocols, methodologies, or models." 45 EPA, Peer Review Handbook, 2015, available at: https://www.epa.gov/sites/production/files/201603/documents/epa_peer_review_handbook_4th_edition.pdf. 46 OMB, Peer Review Bulletin, 2005, available at: https://www.whitehouse.gov/sites/default/files/omb/assets/omb/fedreg/2005/011405_peer.pdf. 47 Some of the refined risk evaluations may lead to costly risk management measures and will thus need to be treated as influential or highly influential scientific assessments. 48 See ACC comments available at: https://www.regulations.gov/document?D=EPA-HQ-OPPT-2015-0805-0012. 19 | P a g e VIII. The Proposed Rule Should Implement a Weight of the Scientific Evidence Approach (WoE) Section 26(i) of the LCSA requires EPA to make decisions using a WoE approach.49 While many groups have described and discussed what is meant by a WoE approach,50 the June 7, 2016, Congressional Record provides a very clear definition.51 This definition, added by senators, is fully consistent with the definition provided by the House of Representatives.52 ACC recommends that, to ensure clarity and consistency in applications, the definition from the Congressional Record be added to the risk evaluation rulemaking as follows: "the term ,,weight of evidence" refers to a systematic review method that uses a pre-established protocol to comprehensively, objectively, transparently, and consistently, identify and evaluate each stream of evidence, including strengths, limitations, and relevance of each study and to integrate evidence as necessary and appropriate based upon strengths, limitations, and relevance." This definition is meant to apply to human and environmental/ecological risk evaluations and should be applied in a fit-for-purpose manner to both screening and refined risk evaluations. As refined risk evaluations must be rigorous to inform potential risk management activities, the WoE standard becomes even more important in this context. However, for all types of risk evaluations, a WoE approach will increase reliability, transparency, clarity, and consistency. The best available science provision at Section 26(h) also applies to both the Section 6(b)(4)(F)(v) and 26(i) WoE requirements, such that the WoE review itself must be based on best available evidence. The rulemaking should make this clear. It should also clearly differentiate between WoE and SoE review, and explain why a SoE review does not meet either the best available evidence or the WoE requirements of the statute. a. Systematic Review is Required Consistent with the definition of WoE, a systematic review is required. This means EPA must, among other steps, provide clear criteria for judging the quality and relevance of all evidence and must then integrate all the evidence based on the identified strengths and limitations and relevance. Systematic review is rigorous. However, once in place, and once the criteria and quality standards are identified and the approach is outlined, the rigor, clarity, and transparency of EPA assessments over time will be greatly improved. There will be tremendous cost, effort, and time savings once fully implemented. ACC recommends that EPA follow the standards for systematic review defined by the Institute of Medicine (IOM).53 The IOM report provides a clear discussion of what is required for a systematic 49 Section 26(i) states, "The Administrator shall make decisions under sections 4, 5, and 6 based on the weight of the scientific evidence." 50 See for example Lutter, R et al., Improving weight of evidence approaches to chemical evaluations; Risk Anal. 2015 Feb; 35(2):186-92. Available at: http://onlinelibrary.wiley.com/wol1/doi/10.1111/risa.12277/full. 51 See Senate Congressional Record, June 7, 2016 at page S3518, available at: https://www.congress.gov/crec/2016/06/07/CREC-2016-06-07-pt1-PgS3511.pdf. 52 See House of Representatives Report 114-175 at page 33, available at: https://www.congress.gov/114/crpt/hrpt176/CRPT114hrpt176.pdf. 53 Institute of Medicine (IOM), Finding what works in Health Care: Standards for Systematic Review, 2011, available at: http://www.nationalacademies.org/hmd/Reports/2011/Finding-What-Works-in-Health-Care-Standards-for-SystematicReviews.aspx. 20 | P a g e review and this approach should be incorporated into the risk evaluation rulemaking. Key elements of a systematic review, which have not changed over time, should include: i. Development of a Protocol The protocol, developed before the risk evaluation begins, defines the methodologies that will be used in the assessment. It is made publicly available before the assessment begins and becomes a living document that can be commented upon and modified as needed. The protocol, arguably the most important part of a systematic review includes: a clear testable question/hypothesis, the planned search strategy (including criteria for inclusion and exclusion of studies), the criteria that will be used for study quality and risk of bias evaluations (including for example consideration of study design and confounders), the plan for integrating/synthesizing evidence using a WoE approach, the plan for dose-response analysis (if necessary), the plan for quantifying and presenting risk findings, and the plan for peer review of the assessment.54 ii. Search Strategy The search strategy (including predefined study inclusion/exclusion criteria, literature sources, search terms, and outreach plan for obtaining stakeholder data) used to identify relevant literature (both negative and positive studies) is well documented and is made available to the public. Any restrictions placed on the literature search or data access are noted and explained.55 iii. Transparency Sufficient data for the critical studies and the models used in the assessment are available to interested external parties so as to enable them to replicate/verify the assessment outcomes and to judge the scientific credibility of the data/models. Confidential business information (CBI) is protected.56 b. A Systematic Review is Not Automatically a WoE Assessment The definition of WoE requires that a systematic review approach be used. A systematic review approach will ensure that EPA"s evaluation is predefined, transparent, and reproducible. The systematic review protocol will define, in advance of conducting the evaluation, the quality criteria and the approach that EPA will use. Unfortunately, the term "systematic review" is often loosely used and many scientists often confuse a systematic literature review with the conduct of a full 54 Finding what works in Health Care: standards for Systematic Review, 2011, available at: http://www.nap.edu/catalog/13059/finding-what-works-in-health-care-standards-for-systematic-reviews; NAS Review of EPA's Integrated Risk Information System (IRIS) Process, 2014, available at: http://www.nap.edu/catalog/18764/review-of-epasintegrated-risk-information-system-iris-process; OHAT systematic review handbook, available at https://ntp.niehs.nih.gov/pubhealth/hat/noms/index-2.html; and EFSA Application of Systematic review, available at: http://www.efsa.europa.eu/en/efsajournal/pub/1637. 55 Ibid; Fenner-Crisp andDellarco, 2016. 56 National Research Council, National Academy of Sciences, Models in Environmental Regulatory Decision Making, 2007, available at: http://www.nap.edu/catalog/11972/models-in-environmental-regulatory-decision-making; Fenner-Crisp and Dellarco, 2016. 21 | P a g e systematic review. A systematic review requires not only a plan for systematically reviewing the literature, but also requires that plans for analysis and data integration be included in the protocol. However, the systematic review process itself does not automatically dictate the approach for analysis and evidence integration. Many systematic reviews, as will be described below, do not necessarily conduct evaluations using a weight of evidence framework for integrating studies based on their strengths, limitations and relevance. This distinction is critically important. EPA, and stakeholders, must be cautious to not conflate WoE and systematic review. The LCSA requires both a systematic review approach and a WoE approach. While both are required, they are separable and the terms should not be confused. The WoE approach mandated by the LCSA is very specific to how EPA will not only weigh information, but how that information should be integrated. This integration step is critically important. c. WoE and Systematic Review for Screening Level Risk Evaluations While it may appear overly rigorous on first glance, both WoE and systematic review approaches can be applied to screening level risk evaluations. This approach is consistent with the approach taken by ECHA where the WoE evaluation is influenced by the amount of information needed and the importance of the decision being taken, as well as the consequences of the decision.57 In screening level evaluations, EPA should use the best available data, based on its strengths, limitations and relevance, as is appropriate (fit-for-purpose) for a screening level evaluation. The protocol for a screening level evaluation should clearly articulate the screening tools, models, data and information that will be used in the evaluation, and should also describe how all the information will be integrated, based on the strengths, limitations and relevance of the data. The systematic review approach will ensure clarity and transparency in the conduct of the screening level evaluations and the WoE approach will ensure that the best screening tools and information are given the most weight. This will ensure that EPA is using the best available science for the screening level evaluations. d. WoE and Systematic Review for Refined Risk Evaluations Similar to the approach described above, both WoE and systematic review approaches must be applied to refined risk evaluations. The key difference will likely be that there exists more data and information, more data streams to evaluate (including mechanistic and mode of action (MOA) information, and more complex considerations of both dose and human relevance, in addition to evaluating study quality and relevance. As these evaluations may inform risk management actions, it is critically important that they are also transparent, reproducible, and rely upon the best available science. Many publications describe tools that exist to evaluate the quality of data, including mechanistic data,58 and provide examples of approaches to integrate the data using a WoE approach.59 57 ECHA, Practical guide How to use alternatives to animal testing to fulfil your information requirements for REACH registration, Version 2.0, 2016, available at: https://echa.europa.eu/documents/10162/13655/practical_guide_how_to_use_alternatives_en.pdf. 58 See for example Lynch H., et al., Systematic Comparison of Study Quality Criteria, Regul Toxicol Pharmacol, 2016, available at: http://www.sciencedirect.com/science/article/pii/S0273230015301525; Greene et al., Challenges in Developing a 22 | P a g e While the LCSA requires EPA to conduct both WoE and systematic reviews, ACC recognizes that while what constitutes a high quality WoE review and a systematic review will not change over time, the approach EPA take may take will likely continue to evolve. ACC encourages EPA to fully engage all stakeholders in the continued development of these approaches. e. Strength of Evidence is Not the Same as WoE A strength of evidence (SoE) approach should not be confused with a WoE approach. Similarly, programs that use SoE approaches should not be considered equivalent to programs that use a WoE approach. EPA should not rely on data from these programs as they do not meet the requirements of the LCSA. SoE is simply not as robust as WoE. Typically SoE approaches have emphasized one or a few studies that report an association between a chemical and a health effect regardless of their quality, replicability, or consistency, and often fail to integrate data gathered from a variety of sources. SoE studies also often ignore negative data. The Office of Health Assessment and Translation (OHAT) at NIEHS conducts systematic reviews and considers the risk of bias of individual studies. However, OHAT does not rely on these findings during integration and OHAT does not conduct a full evaluation of the quality of each individual study. Risk of bias is only a small piece of a full quality evaluation. Similarly concerning, OHAT examines only at the strength of the body of evidence, ignoring the strength of individual studies. Unfortunately, this SoE approach allows OHAT to reach flawed conclusions, e.g., that a body of sub-par studies provides the same level of rigor as a high quality study. This approach is not scientifically defensible. In addition, the OHAT process omits important middle steps relevant to causal inference. The International Agency for Research on Cancer (IARC) also uses a SoE approach, giving more weight to positive studies than negative studies and often paying less attention to the quality and relevance of those individual studies. EPA must employ the Congressionally-mandated WoE approach, not a SoE approach. In implementing the Hazard Communication Standard (HCS), OSHA made clear there are differences between SoE and WoE, acknowledging that WoE goes beyond SoE. In discussing carcinogenicity, OSHA stated that WoE "involves the consideration of other factors, beyond strength of evidence, that influence the likelihood that a chemical may pose a carcinogenic hazard, such as tumor type and background incidence, multisite responses, mode of action, and the comparison of absorption, distribution, metabolism and excretion between test animals and Systematic Review Protocol for Environmental Contaminants In: The Toxicologist: Supplement to Toxicological Sciences, 150 (1), Society of Toxicology, 2016. Abstract no. 2166, available at: https://www.toxicology.org/pubs/docs/Tox/2016Tox.pdf. 59 See for example Lavelle K, et al., Framework for Integrating Human and Animal Data in Chemical Risk Assessment, Regul Toxicol Pharmacol, 2012, available at: http://www.sciencedirect.com/science/article/pii/S0273230011002029; Adami HO et al., Toxicology and epidemiology: improving the science with a framework for combining toxicological and epidemiological evidence to establish causal inference. Toxicol Sci., 2011, available at: http://toxsci.oxfordjournals.org/content/122/2/223.long; Goodman JE., Weight-of-evidence evaluation of short-term ozone exposure and cardiovascular effects. Crit Rev Toxicol., 2014, available at: http://www.tandfonline.com/doi/abs/10.3109/10408444.2014.937854?journalCode=itxc20; Lutter, R et al., Improving weight of evidence approaches to chemical evaluations; Risk Anal., 2015, available at: http://onlinelibrary.wiley.com/wol1/doi/10.1111/risa.12277/full. 23 | P a g e humans."60 This distinction has also been discussed in other publications, where WoE reviews go well beyond the extent of data and information evaluated in a SoE review.61 IX. EPA Should Make Information Available Consistent with Section 26(j) Section 26(j) of the LCSA, while protecting CBI, requires EPA to be transparent not only regarding the data and information the Agency relies upon, but also requires that EPA be transparent about its analyses. The provisions in Section 26(j), in many respects, strengthen the requirements in Sections 6(b)(4)(f), 26(h), and 26(i). In addition, there is a requirement that EPA identify publicly the studies considered, as well as the study results. This requirement will provide increased transparency permitting stakeholders to know which studies that were considered, including those rejected and not used in EPA"s analysis. Coupled with a systematic review, which will clearly state the criteria EPA will use for judging the quality and acceptance of different types of data, this new level of transparency will strengthen confidence in EPA"s chemical assessment program. ACC recommends that all the elements of Section 26(j) be incorporated into the risk evaluation rulemaking. X. EPA Should Use Reasonably Available Information and CBI Consistent with Section 26(k) Section 26(k) of the LCSA requires that EPA takes into account all information, including hazard and exposure information that is reasonably available to the Administrator. Robust study summary data, made available by ECHA, should also be considered by EPA. As part of a systematic review, EPA should release the plans for the literature/data search. Sharing this information, for public comment, early in the risk evaluation process will allow EPA to ensure that useful sources of information and data are not missing when EPA begins a risk evaluation. This step should be a part of the risk evaluation process. EPA must also consider CBI and other information that might have been submitted to EPA for other programs or purposes. ACC encourages EPA to consider CBI as appropriate to inform its risk evaluations; however, EPA must protect that CBI consistent with Section 14 of the LCSA. This includes protecting all CBI contained within material that is otherwise not CBI as provided in Section 14(b)(1). This subsection of 14(b) may have implications for EPA in its use of health and safety studies since EPA will need to continue to protect any CBI that is contained within a health and safety study, e.g., using structurallydescriptive names when chemical identity is claimed confidential. In addition, ACC suspects that a fair amount of information concerning the uses and applications of chemicals, i.e., exposure information, will be CBI. Much of this type of CBI will likely come from processors and/or formulators of chemicals since manufacturers often do not have access to this exposure data. It is essential that EPA work with these downstream users of chemicals to obtain this exposure information and protect their CBI from disclosure while still utilizing the information in risk evaluations. 60 OSHA, Guidance on Data Evaluation for Weight of Evidence Determination: Application to the 2012 Hazard Communication Standard, 2016, at page 12, available at: https://www.osha.gov/weightofevidence/woe_guidance.pdf. 61 See for example McGregor D, et al., Guidance for the classification of carcinogens under the Globally Harmonised System of Classification and Labelling of Chemicals (GHS), Crit Rev in Toxicol, 2010, available at: http://www.tandfonline.com/doi/full/10.3109/10408440903384717. 24 | P a g e A related issue concerns health and safety studies sponsored by companies that are not in the public domain. Some companies will have concerns that certain health and safety studies they have funded will be made public by EPA and then widely available to others who did not make any financial contribution or provide data compensation to the data owner. ACC urges EPA to work with companies to enable them to share studies with the Agency without losing their right to fair and equitable data compensation. EPA should consider making a robust study summary publicly available rather than the full study in such circumstances. If other scientists interested in the actual studies come forward to verify the data and/or quality of the study, etc., those interested scientists should be required to sign a non-disclosure agreement prohibiting personal use of the study for any other purpose. XI. EPA Should Utilize Fit-for-Purpose Exposure Evaluation Tools Recognizing that measured exposure information that is representative, reproducible and reliable is frequently unavailable, EPA will need high quality, fit-for-purpose exposure models to inform both scoping and refined risk evaluations. Many exposure models/tools exist to inform scoping/screening to allow EPA to distinguish uses that present no concern from those that require a targeted higher tier assessment. Consistent with Section 26, models must be used in a manner consistent with "best available science," and they must also be reasonable and consistent with the intended use of the information. Lower tier, very conservative models that are useful for scoping and screening will not be sufficient for the refined risk evaluations necessary to inform risk management decisions. For refined evaluations, EPA should use higher tier exposure tools and refined information, preferably measured data when it exists and is reliable. In addition to EPA"s current tools, ACC recommends that EPA explore the exposure models used in Europe by ECHA, EU member states, and REACH registrants. These tools have been widely used throughout the European Union. Tables 1 and 2 below identify some of the tools available to EPA. Further details on exposure tools are provided in Appendix F. Table 1. EPA Toolbox of Exposure Models62 Screening Assessment Tools Refined Assessment Tools Environmental Worker Consumer Environmental Worker Consumer + + Chemsteer Chemsteer Chemsteer Chemsteer MCCEM E-Fast (v2014) (v3.0) (v3.0) (v3.0) (v3.0) (v2.2) * + E-Fast (v2014) CEM beta (1.3) E-Fast (v2014) WPEM (v3.2) EPI Suite AMEM Sheds-HT beta* (v411) (1990) * denotes models for which validation is not yet complete + denotes models for which data can replace defaults to refine the assessment 62 EPA exposure models from https://www.epa.gov/tsca-screening-tools. 25 | P a g e Table 2. REACH Toolbox of Exposure Models Screening Assessment Tools Refined Assessment Tools Environmental Worker Consumer Environmental Worker Consumer CHESAR CHESAR CHESAR CHESAR ART (v1.5) EGRET (v2) (v3.0) (v3.0) (v3.0) (v3.0)* ECETOC TRA ECETOC TRA ECETOC TRA ECETOC TRA REACT PEST * (v3.1) (v3.1) (v3.1) (v3.1) (2009) StoffenConsExpo PetroRisk BAMA/FEA BAMA/FEA manager (v6) (v4.1) * denotes models for which data can replace defaults to refine the assessment The REACH exposure toolbox includes the Chemical Safety Assessment and Reporting Tool (CHESAR), which incorporates the ECETOC Targeted Risk Assessment (TRA) component for scoping and screening exposures to workers, general population, consumers and the environment. The REACH toolbox also includes BAMA for evaluating aerosol products in indoor environments, and PetroRisk for the evaluation of petroleum products. The TRA tool allows risk evaluation practitioners to conduct a screening level assessment of exposures to workers, consumers, and the environment, all in the same model. Reliable regulatory decision making using a minimum amount of data, at the screening level, is simplified with the TRA. Risks can be assessed in a tiered, integrated approach and communicated in a manner that is relevant and understandable. We encourage EPA to explore the use of this tool, and similar tools to help ensure the efficient, timely, and high quality implementation of the LCSA. Appendix G provides further details on the TRA tool. The toolbox also includes higher tier models like EGRET for solvents, REACT for cleaning products, PEST for plastic additives, ConsExpo for consumer products as other products. For more refined work place exposure, the toolbox includes Stoffenmanager and ART. These REACH models are publicly available and well documented.63 They have been extensively reviewed in ECHA-led conferences, science advisory panels, scientific journals, and many other fora. Importantly, they have been utilized for the registration of over 10,000 substances, establishing their strong track record. ACC encourages EPA to explore the further use of these widely available modeling tools. XII. The Requirements of Sections 6 and 26 Apply to Environmental Risk Evaluations Under the LCSA, EPA will continue to have an obligation to protect the environment. Setting aside the clearly human health specific comments, the framework and methodologies described in these comments can generally apply to both human health and environmental risks. There are certain considerations specific to environmental risk evaluation that merit further discussion. As with human health evaluations, there is an opportunity to improve environmental risk evaluations, and sections 6 and 26 of the LCSA require high quality information and science standards be met. An 63 See for example ECHA Guidance on information requirements and Chemical Safety Assessment Chapter R.16: Environmental exposure assessment Version 3.0, 2016, available at https://echa.europa.eu/documents/10162/13632/information_requirements_r16_en.pdf. 26 | P a g e unreasonable risk determination must be based on hazard and exposure-- it cannot be triggered by volume alone and/or intrinsic properties without assessing behavior in the environment. In a tiered testing paradigm, we suggest that EPA prefer high quality testing data. When not available, EPA should rely on category approaches/ analogues & read-across, and QSAR models. EPA should have a transparent approach for evaluating endpoints/data in each of these categories. Industry scientists often have unique insight and experience with their company"s chemistries and collectively have a large body of knowledge of risk assessment processes globally, including an understanding of potential environmental impacts. ACC encourages EPA to leverage this knowledge and engage early and frequently with industry throughout the risk evaluation process. a. Advancing Models for Environmental Risks Environmental risk evaluation would be greatly improved if there were a larger suite of models that EPA found acceptable. There are many hazard prediction models available of varying quality and reliability, some of which perform better for different types of chemicals. However, the "domain of applicability" restricts which chemicals and chemical families can be modeled reliably in a certain program. Models tailored for specific categories of chemicals may give a more accurate output than more generic models. As such, it would be helpful for EPA to allow a greater range of tailored and externally sourced models into the risk evaluation process, provided the models undergo an assessment for acceptability. ACC would like to work with EPA to develop a more transparent set of acceptance criteria for environmental models. This would permit and accelerate the acceptance of existing and externally sourced models for use in risk evaluation. Stakeholders would benefit from a better understanding of the criteria EPA uses to validate new models, so that users can participate in the acceptance process or eliminate models that are inappropriate. In addition, new models including those specific for certain chemical families could be constructed in accordance with these criteria. These criteria could also be used to generate a running list of pre-approved models. Such a list would provide a degree of flexibility and reliability that will lead to timely and improved risk evaluations. ACC would welcome the opportunity to begin a dialogue with EPA to work together to develop these important criteria for model acceptance. b. Improving Data Sourcing, Generation, and Evaluation There is a wealth of environmental testing data available in REACH. ACC encourages EPA to continue to work with ECHA, OECD, ECETOC, and other stakeholders to move towards standardization of data for the global evaluation and acceptance of environmental studies. Through EPA"s HPV Challenge Program, the OECD Cooperative Chemicals Assessment Programme and REACH, the Klimisch approach has become a widely accepted tool for evaluating the quality of data. More recently, a tool specific for ecotoxicity data, the criteria for reporting and evaluating ecotoxicity data (CRED) has been proposed.64 ACC encourages EPA to adopt an approach such as this through a transparent process that engages all stakeholders in the process. 64 Moermand, CT. et al., CRED: Criteria for reporting and evaluating ecotoxicity data, Environ Toxicol Chem., 2016, available at: http://onlinelibrary.wiley.com/doi/10.1002/etc.3259/abstract. 27 | P a g e ACC also encourages EPA to revisit the state of the science regarding difficult to test substances to allow for the flexibility to provide data that takes the intrinsic properties of the substance into consideration. A better articulation by EPA of the criteria for acceptance of alternative methods for poorly-soluble, difficult to test substances would be extremely helpful and we encourage EPA to look to OECD guidance.65 For instance, the use of activity to describe the degree of saturation achieved by a compound in a given media is particularly useful for poorly soluble lipophilic substances that display a narcosis mode-of-action in aquatic organisms. ACC recommends that EPA find this approach acceptable. In addition to this approach, chemical activities may provide valuable estimates of the proximity of measured concentrations to potentially toxic levels. These values are easy to calculate and allow the comparison of concentration data in various matrices that are of differing units. Finally, ACC encourages EPA to adopt new analytical approaches, for polymers and UVCBs, regarding water solubility. ACC members have a wealth of expertise in this space, including experiences from working with Canadian and Korean governments. ACC would welcome the opportunity to discuss these approaches with EPA. c. Persistent, Bioaccumulative and Toxic (PBT) Substances PBT substances are substances that resist degradation in the environment and in organisms, leading to progressively higher and unpredictable concentrations in a food web and therefore may pose risks to top predators and humans. When assessing a substance for potential PBT properties, EPA must go beyond the numerical screening criteria used today that screen only for the potential to behave as PBTs in the environment. Recent publications have discussed this approach in the context of PBTs.66 Relevant reliable data should not be excluded simply because screening criteria are met. For example, a bioconcentration factor may indicate that a substance could bioconcentrate in a fish but if the substance is not found in water, cannot reach levels in the fish that can cause toxicity and will not biomagnify in food webs due to metabolism. All of this data should be used to assess if that substance is actually bioaccumulative in the environment. In addition to using all data to assess these endpoints, it is important to evaluate each study for its quality and relevance to the endpoint being assessed, taking into consideration the physical chemical properties of the substance. XIII. EPA Should Leverage International and Inter-Agency Cooperation Section X, above, discusses the need for EPA to ensure that information from other offices within EPA is considered to inform risk evaluations. However, the universe from which EPA could obtain useful information is much larger. EPA should leverage data and information reasonably available from other jurisdictions where that data and information has applicability in the United States and is of sufficient 65 See for example OECD Guidance Document on Aquatic Toxicity Testing of Difficult Substances and Mixtures, available at: http://www.oecd-ilibrary.org/environment/guidance-document-on-aquatic-toxicity-testing-of-difficult-substances-andmixtures_9789264078406-en. 66 See for example, The Origin and Evolution of Assessment Criteria for Persistent, Bioaccumulative and Toxic (PBT) chemicals and Persistent Organic Pollutants (POPs), M. Matthies et al, Environ. Sci.: Processes Impacts, 2016, DOI 10.1039/C6EM00311G, and Comparing Laboratory and Field Measured Bioaccumulation Endpoints, Burkhard et al, IEAM, Vol 8, Number 1, 2011. 28 | P a g e scientific quality to meet the science standards required under the LCSA. For example, there is a significant amount of information available, including robust study summaries, associated with submissions to ECHA for REACH chemical evaluations.67 In addition, EPA"s work with Canada under the Regulatory Cooperation Council (RCC) has provided EPA with data and information relevant to several case studies on chemicals that are on EPA"s 2014 Updated Work Plan. That information and the learnings from that work should be leveraged by EPA in its risk evaluation work under the LCSA. Similarly, EPA should not be compelled to further evaluate a chemical and its conditions of use if another regulatory authority (either in the US or elsewhere) has already conducted an evaluation, provided the uses and exposures are comparable to those in the United States and the quality of the evaluation meets the scientific standards required under the LCSA. EPA has relied on assessments completed by other jurisdictions and agencies in the past and there is no reason it should not continue to do so in appropriate circumstances where the scientific quality meets the standards dictated by the LCSA and conditions of use are relevant to EPA"s risk evaluation. At a minimum, EPA should coordinate with other federal agencies with similar/sometimes overlapping/relevant jurisdictions as contemplated by Section 9 of the LCSA. XIV. Incorporating High Throughput Tools and Alternative Methods To improve efficiency, particularly for screening level risk evaluations, EPA will need to adopt alternative methods that allow for the evaluation of chemicals that may not have robust testing data available. Read Across and QSAR modeling are important toxicological tools that can be used to assess hazards and risks of a substance without conducting additional animal toxicity tests. 68 Read Across uses relevant information from analogous substances to predict a specific toxicity endpoint for the target substance.69 EPA, OECD and industry have gained considerable experience using read across approaches under both EPA"s HPV Challenge Program and the OECD Cooperative Chemicals Assessment Programme. Both EPA and OECD have issued guidance on methods for forming categories for use in read across.70 More recently, ECHA has developed the Read Across Assessment Framework (RAAF) which should also be considered by OPPT.71 67 Non-confidential REACH data is made easily available through AMBIT, which can be accessed at: http://ceficlri.org/lri_toolbox/ambit/. 68 See for instance http://ehp.niehs.nih.gov/1104666/ and https://echa.europa.eu/documents/10162/13655/pg_avoid_animal_testing_en.pdf. 69 Patlewicz et al., Read-Across Approaches - Misconceptions, Promises and Challenges Ahead, 2014, available at http://altweb.jhsph.edu/altex/31_4/FFTPatlewicz.pdf ("Read Across from Analogs/Categories - ,,Read across" is a technique of filling data gaps. To ,,read across" is to apply data from a tested chemical for a particular property or effect (cancer, reproductive toxicity, etc.) to a similar untested chemical. The read across technique is often applied within groups of similar chemicals assembled for assessment using either analog approach (grouping based on a very limited number of chemicals) or category approach (grouping based on a larger number of chemicals). In an analog/ category approach, not every chemical needs to be tested for every endpoint.") 70 See EPA, Development of Chemical Categories in the HPV Challenge Program, available at: https://web.archive.org/web/20080829212006/http://www.epa.gov/chemrtk/pubs/general/categuid.htm; OECD, Guidance on Grouping of Chemicals, Second Edition, available at: http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=env/jm/mono(2014)4&doclanguage=en 71 See https://echa.europa.eu/documents/10162/13628/raaf_en.pdf. 29 | P a g e Historically, Read Across has relied upon traditional toxicity datasets; however, advanced approaches for biological profiling, such as high throughput screening and high content profiling, hold considerable promise for improving the scientific basis for developing categories, analyzing analogues and supporting quantitative (e.g., relative potency) Read Across.72 Advanced approaches, such as high throughput screening (e.g. ToxCast) and high content profiling (e.g., transcriptomics), can also be used to determine initial molecular events and early cellular responses (early key events) as part of an MOA analysis. These technologies hold considerable promise in establishing confidence that a given substance produces initial molecular events and early key events that lead to an adverse effect by a specific MOA and, importantly, such evidence can form the foundation for the scientific justification for selecting dose-response extrapolation methods to be used for human health risk evaluations. In addition, the use of data from in vitro studies can be calibrated for relevance using exposure data and quantitative in vitro to in vivo extrapolation methods. In developing the path forward for guidance and regulations on scoping and risk evaluations, OPPT is encouraged to include plans to work closely with EPA's National Center for Computational Toxicology and stakeholders to establish confidence in these advanced biological technologies for use in advanced biological profiling technologies for use in Read Across, adverse outcome pathway (AOP) development, and MOA determinations. XV. Stakeholders and EPA Must Be Held to the Same High Standard Sections 6 and 26 set high quality standards for the conduct of risk evaluations that must be incorporated into the rulemaking. These quality requirements will increase the rigor and credibility of EPA"s chemical evaluation program for decades to come. We recommend that these same high quality standards be incorporated into the guidance for stakeholders to inform the submission of draft risk evaluations which shall be considered by the Administrator as required in Section 26(l)(5). The high quality standards required by the LCSA should apply to everyone. The new statute demands it and ACC and its member companies stand ready to be held to these same high quality standards. 72 See for example http://www.ncbi.nlm.nih.gov/pubmed/27026708. 30 | P a g e Appendix A: Principles for Improving Chemical Hazard and Risk Assessment Communication Review and Accountability ACC's Principles for Improving Chemical Hazard and Risk Assessments Assessments should focus on understanding the inherent properties of substances in order to determine the likelihood of harm from a specific exposure. The public, businesses and regulators look to government assessments for reliable information about the potential hazards and risks associated with chemicals. Identify Key Science Issues Prior to Initiation of Assessment 0 Discuss the purpose, scope and technical approaches 0 Engage stakeholders Apply Objective Criteria 0 Develop and apply consistent criteria for selecting and evaluating a study, before an assessment begins Evaluate all studies to determine their quality, relevance and reliability nsu re Assessments a re Transparent Disclose key information and assumptions used to develop assessments and reach conclusions . Make materials, including important data sets, publicly available Conduct Scientific Peer Review by Independent Experts 0 Ensure peer reviewers are fully independent from the program office issuing the assessment 0 Evaluate peer review panels for conflicts of interest; ensure panels contain a balance of perspectives and appropriate technical expertise Use Modern Science and Tools I Use relevant data 0 Consider how chemicals act in the body 0 Evaluate chemicals at relevant exposure levels Integrate Evidence . Give the greatest weight to information from the highest-quality and most?relevant studies 0 Transparently and objectively integrate evidence to make realistic determinations of hazards and risks; consider all types of evidence Characterize Hazards and Risks Fully and Accurately - Present hazards and risks in an easy?to- understand manner to stakeholders and risk managers Present a range of plausible values, including central estimates when going beyond a screening level assessment Improve Accountability 0 Use an independent accountability procedure to verifythat revised assessments are accurate and responsive to scientific and peer review RESULT: Public Trust in High?Quality Risk Assessment 31|Page Appendix B: Improving Hazard Assessment ACC recommends that these important steps in a hazard assessment be incorporated into EPA"s rulemaking for risk evaluation. This will ensure that the conduct/process of developing refined risk evaluations is consistent with the high quality demanded by Sections 6 and 26 of the LCSA. Critical Steps for Hazard Assessment: 1. A pre-defined established weight-of-evidence approach, addressing causal relationships, is applied in a systematic manner to integrate, weigh the lines of relevant evidence, and effectively use all relevant information. This includes ensuring that both positive and negative studies are weighed objectively and that the highest quality and most relevant studies are given the most weight/consideration. Judgments and choices should all be transparently presented. A "strength of evidence" approach is not acceptable.73 Section VIII of these comments discusses the weight of evidence (WoE) concept in more detail. 2. EPA should include a robust discussion of key lines of evidence and inherent uncertainties, alternative interpretations, other issues that may have prompted debate, and how these issues are addressed.74 3. EPA should identify, explain, and discuss the level of adversity of the chosen endpoint(s).75 4. Biologically plausible mode of action (MoA) information should be considered and fully incorporated. The MoA analysis should include a consideration of category analogs as a complement to chemical specific data, and existing knowledge must be leveraged on already established MoAs similar to the substance of interest.76 5. EPA should provide a discussion of whether the key events within the MoA would progress to an adverse effect relative to concentration/dose and anticipated human exposure (duration/magnitude/route), and life stage.77 73 NAS, 2014 ; Fenner-Crisp and Dellarco, 2016. Ibid. 75 EPA, Risk Assessment for Noncancer Effects, available at: https://www.epa.gov/fera/risk-assessment-noncancer-effects; EPA, A Review of the Reference Dose and Reference Concentration Processes, 2002, available at: https://www.epa.gov/sites/production/files/2014-12/documents/rfd-final.pdf. 76 EPA, Guidelines for Carcinogen Risk Assessment, 2005, available at: https://www3.epa.gov/airtoxics/cancer_guidelines_final_3-25-05.pdf; Fenner-Crisp and Dellarco, 2016. 77 EPA Guidelines for Carcinogen Risk Assessment, 2005; IPCS Mode of Action/Human Relevance Framework, available at: http://www.who.int/ipcs/methods/harmonization/areas/cancer/en/; Fenner-Crisp and Dellarco, 2016. 74 32 | P a g e Appendix C: Improving Dose Response Assessment ACC recommends that these important steps in a dose response assessment be incorporated into EPA"s rulemaking for risk evaluation. This will ensure that the conduct/process of developing refined risk evaluations is consistent with the high quality demanded by Sections 6 and 26 of the LCSA. Critical Steps for Dose Response Assessment: 1. The dose responses should be plotted for all relevant non-cancer endpoints of concern and a distribution of hazard values or points of departure (POD) should be provided for all relevant endpoints. The selection of the hazard values must be well justified and supported by the overall database.78 2. When biologically plausible, potential carcinogenic effects should be modeled and presented using nonlinear approaches in addition to linear modeling approaches.79 3. A non-linear, point of departure modeling approach should be used for non-genotoxic carcinogens. For genotoxic carcinogens acting through MOAs that are clearly understood to be threshold events (e.g., clastogenesis induced by DNA-DNA and/or DNA-protein crosslinks such as binding to spindle apparatus), a non-linear model should be used.80 4. EPA must take into consideration natural background levels as well as endogenous human production of compounds when evaluating dose-response. Hazard values below background should not drive risk management determinations.81 5. The endpoints used in the dose-response assessment should be those associated with adverse responses in humans, biologically plausible in humans, and derived from studies of high quality and relevance. Bradford Hill Considerations should be used to evaluate critical endpoints.82 6. The nature of responses (e.g., biochemical, morphological, physiological or functional change, severity of the effect, reversibility) and their dose-responses (e.g., steepness or shallowness of dose-response curve, dose spacing between NOAEL and LOAEL) should be clearly described.83 7. Consistent with the level of complexity needed and if data support modeling, multiple approaches should be carried forward in the analysis and a justification must be provided for model selection.84 78 Ibid. EPA, Guidelines for Carcinogen Risk Assessment, 2005; see also NAS, Health Risks from Dioxins and Related Compounds: Evaluation of the EPA Reassesment, 2006, available at: http://www.nap.edu/catalog/11688/health-risks-from-dioxin-andrelated-compounds-evaluation-of-the. 80 EPA, Guidelines for Carcinogen Risk Assessment, 2005; Preston and Williams, DNA-reactive carcinogens: mode of action and human cancer hazard, Crit Rev Toxicol., 2005 Oct-Nov; 35(8-9):673-83, available at: http://www.ncbi.nlm.nih.gov/pubmed/16417034; Andersen et al., Dose-response approaches for nuclear receptor-mediated modes of action for liver carcinogenicity: Results of a workshop, Crit Rev Toxicol. 2014 Jan;44(1):50-63, available at: http://www.ncbi.nlm.nih.gov/pubmed/24083384 81 NAS, Science and Decisions, 2009, available at: http://www.nap.edu/catalog/12209/science-and-decisions-advancing-riskassessment; Fenner-Crisp and Dellarco, 2016. 82 EPA, Guidelines for Carcinogen Risk Assessment,; EPA, A Review of the Reference Dose and Reference Concentration Processes, 2002. 83 EPA, Risk Assessment for Noncancer Effects; Fenner-Crisp and Dellarco, 2016. 79 33 | P a g e 8. EPA should use reliable data in lieu of default assumptions or models as a preferred approach. Any default assumptions should be clearly identified and the rationale for each must be explained including describing the impact of the default on the assessment"s conclusions.85 9. Consistent with the level of complexity needed, suitable toxicokinetic and toxicodynamic data must be used to derive more refined dose response estimates. If available, quantitative dose-response information regarding key events within a MoA should also be incorporated into the modeling.86 84 EPA, Benchmark Dose Technical Guidance, 2012, available at: https://www.epa.gov/sites/production/files/201501/documents/benchmark_dose_guidance.pdf; Fenner-Crisp and Dellarco, 2016. 85 EPA, Guidelines for Carcinogen Risk Assessment, 2005; OMB/OSTP memorandum "Updated Principles for Risk Analysis," 2007; Fenner-Crisp and Dellarco, 2016. 86 EPA, OPP Guidance on Tiered approach to Tolerance Assessment, available at: xx; EPA, Guidelines for Carcinogen Risk Assessment, 2005; Fenner-Crisp and Dellarco, 2016. 34 | P a g e Appendix D: Improving Risk Characterization ACC has identified key aspects of risk characterization that are critical to ensuring that risk evaluations are consistent with the science standards in the LCSA. This approach is fully consistent with EPA"s Risk Characterization Handbook, last updated in 2000. ACC recommends that each of these key aspects be incorporated into the risk evaluation rulemaking. This will ensure that the conduct/process of developing refined risk evaluations is consistent with the high quality demanded by Sections 6 and 26 of the LCSA. Key Aspects of Risk Characterization: 1. Risk characterization chapters should be written for both technical and non-technical audiences and be clear and understandable in describing the purpose, objectives, scope, and main findings.87 2. Consistent with the scope and context, all potential hazards/risks should be presented for the populations and exposure scenarios of interest.88 3. A Margin of Exposure (MOE) approach should be used to present findings. Expected and central estimates should be presented, as well as appropriate upper and lower bound values.89 4. This section should incorporate principles of Transparency, Clarity, Consistency and Reasonableness.90 5. EPA must ensure that the analysis presented is consistent with data that meet the relevance and quality criteria and minimizes biases related to study design, data selection, data interpretation, model choices, and conclusions.91 6. Scientific facts and science policy choices must be distinguished.92 7. Confidence in conclusions/risk values should be placed clearly in the context of certainties and uncertainties, and the reasoning for use of and impact of defaults on conclusions must be explained.93 8. Alternative judgments, hypotheses and models must be presented along with support explaining these alternatives. If the assessment includes only a worst-case scenario, an explanation and discussion of uncertainties must be provided.94 9. Significant data needs are clearly identified. There is discussion of the potential impact such data might have on the assessment (i.e., value of information).95 87 EPA, Risk Characterization Handbook, 2000; Fenner-Crisp and Dellarco, 2016. Ibid. 89 EPA, Risk Characterization Handbook. 90 Ibid; Fenner-Crisp and Dellarco, 2016. 91 Ibid. 92 Ibid. 93 Ibid. 94 National Research Council, National Academy of Sciences, Understanding Risk: Informing Decisions in a Democratic Society 99-100, 1996, available at: http://www.nap.edu/catalog/5138/understanding-risk-informing-decisions-in-a-democraticsociety; Fenner-Crisp and Dellarco, 2016. 95 EPA, Risk Characterization Handbook, 2000; Fenner-Crisp and Dellarco, 2016. 88 35 | P a g e 10. The assessment should compare predicted or modeled health outcomes in relevant populations with actual outcomes.96 11. The rational for the use of defaults and models, in lieu of data, is clearly explained and justified.97 12. Comparisons are provided to other assessments that have evaluated the same risks. A discussion of any conflicting results should be provided.98 96 EPA, Risk Characterization Handbook,. Ibid. 98 Ibid. 97 36 | P a g e Appendix E: Ensuring Robust Peer Review ACC has identified key aspects of peer review that are critical to ensuring that risk evaluations, particularly refined risk evaluations, are consistent with the science standards in the LCSA. This approach is fully consistent with EPA"s Peer Review Guidance, the best practices of EPA"s Science Advisory Board (SAB) as well as the recommendations from the Office of Management and Budget. ACC recommends that each of these key aspects be incorporated into the risk evaluation rulemaking. This will ensure that the conduct/process of developing refined risk evaluations is consistent with the high quality demanded by Sections 6 and 26 of the LCSA. Key Aspects of Peer Review: 1. A documented process for peer review that matches the purpose/scope and potential impact of the assessment must be provided for public comment before the assessment begins.99 2. Panel composition shall be consistent with best practices and ensure sufficient knowledge, expertise, and depth. Biases/perspectives shall be identified and balanced. Conflicts of interest shall be identified and disclosed.100 3. All draft materials should be made available to peer reviewers and the public at the same time, allowing adequate time for review and comment.101 4. Peer reviewers should receive public comments in advance for adequate consideration before the peer review meeting is conducted.102 5. At least 45 days should be provided for public comment and review of technical information.103 6. At least one public peer review meeting will be held.104 7. 8. There will be reasonable opportunity and adequate time for public comments to be presented at the public peer review meeting. There is an opportunity for peer reviewers to engage with public commenters on the key technical issues they put forward.105 If peer reviewers did not reach consensus, a minority opinion/report will be provided.106 9. Peer reviewers, in their written report, provide responses to substantive public comments.107 99 EPA, Peer Review Handbook, 2015; OMB, Peer Review Bulletin, 2005; Fenner-Crisp and Dellarco, 2016. EPA, Peer Review Handbook, 2015; OMB, Peer Review Bulletin, 2005. 101 Ibid; Fenner-Crisp and Dellarco, 2016. 102 Ibid. 103 This is consistent with EPA"s Work Plan chemical past practices. 104 EPA, Peer Review Handbook, 2015; OMB, Peer Review Bulletin, 2005. 105 Ibid; Fenner-Crisp and Dellarco, 2016. 106 EPA, Serving on the EPA Science Advisory Board: A Handbook for Members and Consultants, 2012, available at: https://yosemite.epa.gov/sab/sabproduct.nsf/WebBOARD/Serving%20on%20the%20EPA%20Science%20Advisory%20Board: %20A%20Handbook%20for%20Members%20and%20Consultants/$File/Serving%20on%20the%20EPA%20Science%20Advi sory%20Board%20SABSO-12-001.pdf;Fenner-Crisp and Dellarco, 2016. 107 H.R. 1029, EPA Science Advisory Board Reform Act of 2015, available at: https://www.congress.gov/bill/114thcongress/house-bill/1029/actions. 100 37 | P a g e 10. Public and peer review comments are addressed and a response to comments document is released when the final assessment is released.108 108 EPA, Peer Review Handbook, 2015; OMB, Peer Review Bulletin, 2005; Fenner-Crisp and Dellarco, 2016. 38 | P a g e Appendix F: Exposure Modeling Tools Below we provide information and links to exposure modeling tools used by ECHA, and others, to help inform robust exposure evaluations. REACH Tools: Screening/Scoping: ECETOC TRA (v3.1 2014): Tier 1 screening level tool to calculate the risk of exposure to chemicals to workers, consumers and the environment. Used in REACH submissions. Could be used for Scoping and Screening. http://www.ecetoc.org/tools/targeted-risk-assessment-tra/ CHESAR (v3.0): Tier 1 screening level tool developed by ECHA to carry out REACH safety assessments. Could be used for Scoping and Screening. https://chesar.echa.europa.eu/ ? Incorporates: o TRA o MEASE (for catalysts) o RiskofDerm o EMKG Expo Tool BAMA/FEA: Used to generate predicted concentration of aerosol components within workplaces following suitable time interval after spraying. http://www.aerosol.org/regulatory-policy-affairs/productsafety/indoor-air-quality PetroRisk: Designed to evaluate environmental exposure and ecological risks at both local and regional scales for a wide range of petroleum products from naphtha (gasoline), kerosene, gas oils, to heavy fuel and lubricant oils as well as hydrocarbon-based solvents. The spreadsheet tool can evaluate risks associated with different stages in the product life cycle. https://www.concawe.eu/reach/petrorisk Higher levels Tools: EGRET (v2): Tier 1.5 tool developed by the European Solvent Industry Group for evaluating consumer exposure to solvents in REACH submissions. Could be used for Refined Risk Evaluation. http://www.esig.org/en/regulatory-information/reach/ges-library/consumer-gess ART (v1.5): Tier 1.5 tool developed to evaluate worker exposure to inhalable mists, dusts and vapors. Could be used for Refined Risk Evaluation. https://www.advancedreachtool.com/ PEST : Tool developed to for refined assessments of exposure to plastic additives. Could be used for Refined Risk Evaluation. http://www.plasticseurope.org/plastics-sustainability-14017/consumerprotection/reach.aspx REACT (2009): Tier 1.5 tool to estimate systemic consumer exposures to substances that are present in cleaning preparations. Could be used for Refined Risk Evaluation. https://www.aise.eu/ouractivities/product-safety-and-innovation/reach/consumer-safety-exposure-assessment.aspx ConsExpo (v4.1): Comprehensive Consumer Exposure model developed by RIVM (NE) in 1996. Version 4.0 has separate modules for inhalation, dermal and oral exposure. In English. Based on transparent calculation. Could be used when have consumer use that is not included in the E-FAST model. 39 | P a g e Has been used by Canadian Government in their CEPA program and EU REACH. Could be used for Screening or Refined Risk Evaluation. http://www.rivm.nl/en/Topics/C/ConsExpo Stoffenmanager (6): Tier 1.5 recognized by the Dutch Labor Inspectorate and by ECHA as a reliable tool for use in the assessment of exposure situations via inhalation as part of the Risk Inventory and Evaluation (RI&E). In addition, Stoffenmanager is also capable of assessing the risks of exposure via the skin. https://stoffenmanager.nl/Default.aspx Additional Tools for Consideration: Screening Assessments: AIHA models: IH SkinPerm v1.2 and IH MOD are tools developed and widely accepted to estimate worker exposure. https://www.aiha.org/get-involved/VolunteerGroups/Pages/Exposure-AssessmentStrategies-Committee.aspx CalTOX (v2.3): Relates the concentration of a chemical in soil to the risk of an adverse health effect for a person living or working on or near the contaminated soil. https://www.dtsc.ca.gov/AssessingRisk/caltox.cfm USETOX (v2.0): Based on scientific consensus for characterizing human and ecotoxicological impacts of chemicals in life cycle impact assessment. The main output includes a database of recommended and interim characterization factors including environmental fate, exposure, and effect parameters for human toxicity and ecotoxicity. http://www.usetox.org/ RAIDAR (v2.0): Model to screen and prioritize large numbers of chemicals based on hazard, exposure and risk assessment objectives for more comprehensive, higher-tiered assessments. http://www.arnotresearch.com/index_download1.html#!/page_Downloads Refined Assessment: IAQX (v1.1): An indoor air quality (IAQ) model that complements and supplements existing IAQ simulation programs. Inhalation only. IAQX is a Tier 2 model for advanced users who have experience with exposure estimation, pollution control, risk assessment, and risk management. https://www.epa.gov/air-research/simulation-tool-kit-indoor-air-quality-and-inhalation-exposure-iaqx 40 | P a g e Appendix G: Additional Information on the ECETOC TRA The slides attached below are from a presentation developed by Dr. Chris Money with Cynara Consulting. Please note: An explanation of many of the REACH related terms used in the slide below can be found in a Glossary on the UK health and Safety executive website: http://www.hse.gov.uk/reach/definitions.htm. 41 | P a g e Driver for the TRA - established a Targeted Risk Assessment task force to highlight that Effective chemicals regulation needs to be informed by risk Exposure needs to be adequately accounted for as a key consideration when assessing and managing risk The ef?cient acquisition and application of hazard and exposure data can only be brought about through the application oftiered. risk-informed processes Reliable screening level (Tier 1} risk assessments can be undertaken by 'non-experts' Appropriate decisions can be reliably made in the absence of a full 'base set' of hazard data ECETOC is a recognised and astabiis'ted Eti scienti?c organisation. itis funded by over 50 companies with an intarestin cnen'iicaisafaty 3 The Genesis of the TRA - Exposure AssessmentGuidelines recommend completing exposure assessments iteratively using a tiered approach to ?strike a balance between the costs ofadding detail and re?nement to an assessment and the bene?ts associated with that additional re?nement? - When conducting a tiered exposure assessment. after each iteration. the question is asked. - Is this level of detail or degree of con?dence good enough to achieve the purpose of the assessment? - lfthe answer is no. successive iterations continue until the answer is af?rmative. new input data are generated. or. as is the case for many assessments.the available date. time. or resources are depleted. - 3 EPA 1932} Guidelines for Exposure Assessment Screening vs Refined Assessments Screening Re?ned . Readily available data Site- or scenario- Inputs Conservativetdefault spec-i?cdata assumptions Realistic assumptions Point estimates Distributions ofdata - Simple modelsand Complexmodels and T0545 equations equations - Deterministicapproach Deterministicor probabilisticapproach . Gangewatiue egtimate of exposure estimate RESUIE - Usefulfor prioritization 33:32:32: better - IGreateruncertaintv characterised - variabilitvnotgenerallv considered 42| The Challenge Can the principles ciftiered rislt assessment be imprcived such that acceptable decisicuns can rciutineli,r be achieved earlier in the prbcess Can same bfthe advantages bf re?ned teels be transferred tb screening medels?r Hevv can the applicatien bf re?ned mbdels be better targeted (er streamlined} Hcivv tb capitalize frcim the speed and casts level assessments vvithdut being ccinstrained bv their uncertainty.r {in abscilute terms} and inabilityr to characterize variabilitv??I Outline was the TEA activitj-.- initiated by; Eurapean strj-.- Hcivv has the fcicus and ccintent cf the TRA changed civer time??J And vvbv'? What are ccinsidered td be the kev attributes?? Where has the TRA impacted Hcivv can the principles be applied tb the prbcess of exposure assessment?? Core Aims of the TRA v1 {2002} Tb FOCUS resburces cin general substance prbductidn and use scenaricis that ccinstitute a likelv ccincern fcir humans or the envircinment Tb ensure that all decisicins are based upcin RISK and account fer the relevant infcirmaticin that might be expected tci be available Tb but maintain the scienti?c integritv of the RA precess 43 I I?u ?l OtherAims Behind the Approach - To develop an userfriendly {product that enables manufacturers and users of chemicals to readily evaluate and identify chemical SHE rislts To clearly demonstrate the scientific integrity of the proposals - To deliver consistency 1.Ivith the expectations of European regulations is. H35 and IChemical?tgents Directive - Align with other 'accepted' toolsfconcepts for exposure?driven risk assessmentto facilitate harmonization across domains - To transparently demonstratethe utility and integrity of the concepts via a '-.veb-based tool Underlying Approach of the TRA Tiering and Targeting Uses of {louse-m Demonstrateadeqimyof ris or inmlenent further rislr. redidioneg Fm. based on s'mple substitution rules and readily available data I?p Rhtocusesontherisks . .. . fromthepnmaryuses 31-53"qu3 ofthe substance La. 2 Targetedeiforthose scenarios of concern (30111511141115 if consistentvvitit EU an. Immediate Consent '5 CE TVA-R: 0'3 pri ei pies TRA version 1 {zoos?4} I Scope of Tool I Uses of Demonstrateadeqimyot ris koontrols or implement further rislr. remdion e.g. Fi?. based on s'mple substitution rules and readily available data 1 I Rhtocusesontherisks . .. . fromthepnmaryuses 31-53"qu3 ofthe substance I I Targeted HAforthose scenarios of concern of No EtaJ es cons i stentvvitit EU an. Immediate Consent CE F?s?El 0?3 pri ei pies TRA and the Final REACH Text . Drafts of the REACH Regulation began to appear in 2004 The text differed signi?cantly,r from the White Paper in its 'acceptance' oftiering (for hazard}: the need to demonstrate and communicate conditions of'safe use' ['esposure scenarios'}; and phase-in dates Eiut the text offered no practical solutions for hoir.r the basic legal requirements might be accomplished Industry,r supported the application of the TRA in order to consistently and develop REACH RAs And SEWE as a platform for testing emerging ideas being suggested in Technical Guidance' e_g_ Lise Descriptors TRA version 1 {2002-6} I Scope of Tool I Of Concern Demonstrateademmyof ris koontrols or intplemsnt further rislr. redidion e.g. an. based on s'mple substitution rules and readily available data ?El? 1 I I?p. Fi?. focuses on the risks .. . fromthepnntawuses ofthe substance LE. 1 I Targeted FtAforthose scenarios of concern Commons of Na ?11: os cons i stentwith EU as Immediate Concern '3 CE DEE pri ci pies TRA version 2 (see? onwards} Scope of TEA Tool Demonstratsadeqimjrot ris koontrols or implement further rislr. recliction sag. substitution Tier 1 Fmtocuses on the risks 1 from the primanr uses of the substance I . Targeted FtAforthose I scenarios of concern EIPQSIE consistentwith EU SEEN-LEEDS principles The significantly,r updated TRA consists of 2 elements - The general philosophy oftiering and targeting rislt assessments - Uptimia?ngthe supporting TFta exposure toolfor application atthe Tiert levef 45| Impact of the TRA The has been applied in of REACH CSAs substance registrations The is used as the basis of ISSi'Jttool {Chesar} The EU regulatory communityis applyingthe modelto determinethe acceptabilityof risksto human health and related data needs chosar The model demonstrates that comprehensive chemical rislc assessments can be quickly, efficiently and reliably undertaken Comprehensive CSAs possible in lessthan 2 hours Covering health and the environmentacrosethe substance life cycle Without requiring acceseto specialist consultants The has catalyzed a number of initiatives aimed at further improving the relevance, sensitivity, efficiency and consistency of exposurelrislc assessments '5 TRA as a Catalyst for Targeted Action Nature of life cycle Effectiveness of Use Descriptors Tier 1 exposure controls +LJserl-1aps templates . simulations Widespread use oftl'le TRAmodelbeusedas a leverforobtainingan improved Nature of controls ??dem?t?ndng?f Gaps in during use Em?suresingenem'? understanding of exposure +Derma|exposures +Aeroso|s +Fume condensates - Related initiatives aimto improve ef?ciency. relevance and understandability ofRA Enhancing the Relevance Key Input Purpose Status Elements Descriptors Coding rrecl'arisn' forerablirg the Recertly revise-:1 ar-J Lpzlatezl. consistertclescription o" th'ar ar-zl Mara-gel by EC HA. Stare: at er viror mertalexpoeres SEE IS- US El F35 Bose: cescriptior c" wl'ere [arc l'crn'} Exist "or rrost rrajor sL pply sL ces are se: witf ir sL pply cl'air s. h?arage: by ir-: stry cl'air sectors SDERCS Eescriptior of tl'e rature o" Exist "or many n'ajor er iror mer tel en' issiors for oer'irel processes. Mar-age: by in-:l Lsuy ir-J strial processes sectors. All; witl' DEC EC SEEDS Descriptior of tie ratLre cu" ccrerrer Exist for rrary rrajor corerrer expc-sL res "or -:e"ire: ses of cor sL rrer pro: ct roL ps. h?ara-ge: by prc:Lcts [arc articles} sectors. Eescriptior of tl'e rature o" typiml Exist "or principle cl'airs. marker expoere cortrols for-Jefirezl Mar-age: by sectors. I'l'l are professional processes. Carr plernert GESs [see below} 46| Conservatism of the TRA - How conservative should a model be to successfully,r operate at the Tier 1 level? - How far should it re?ect the 'worst case'?r -- 1it'li'orlters . if: - More realisticestimatesthan previous ELJ 2: preferredworxermodel{EASE} .j - Intendedto alignwith T?Emi-i: for responsible use Consumers - Estimates based on reasonable worst case use conditions: iforeseeable use' - Estimates affected bvvariancein ELJ consumer protection approaches Environment - More realistic estimatesthan EU ELJSES model - help deliverimproved estimates an." luau-nun- in?: 5 Observations The TRA has evolved over time Recent changes have been associated with wide stakeholder engagement {in Europe} Maintaining high pro?le exposure tools is resource demanding Testing: revisions: help and support functions: language. etc. The TRA historv charts a series of concepts many,r of which are also relevant to exposure assessment in the US E_g_ an effective Tier 1: nature of integration ofhazard and exposure in FHA: de?nitions of use: importance of conditions of use: abilitvtofocus on usesfcontrols of concern I ul TRA Outreach 55H. :35- CH- I I EUEE I scar-3:; 5.3512}: sis Was-1 I W5- I w; I i016- REQUlatDrYa-nd Hort-cw Tern 2 it": ?c REA-2H Stakeholder Review come in canal: an? 511,05 I Egan-.13.: I THE Activities- ?rm-2:59. Norber- Jill-:3: rue EE-?Jeaco?n I '3?33' . ?m 3:93.51: I ammo Ease-c Won I I Berna Iva canon mud-1 I 2933' CI 2.3.- 5 h. 5355? Scienti?c Review I andln ut 53"5 Les?sat: 131:1 mm "me-'me? Jon?Inc Peerxrticies new. salts-a5: 15?, - Maximising opportunitiesforscienti?c and stakeholder review andfeedbacx 2'3 47| Improving Process Efficiencies How can many,r substances and scenarios be dealt with e?icientlv?r Useful aspects have proven to be: - An inter-connected worl-cer. consumer and environmental model single data entrv for substanoelD and pchem - Elatch mode capabilities all scenarios can be evaluated simultaneouslg.r and - Common framework for describing exposures ef?cienciesin consistenov. deliver relevant user exposurecontrols. facilitate common inputs. help with CEII. enable sensitivityr analvsis on control options - Standard defaults transparenov and ease of use; modi?cations allowed within limits and subjectto justification iv How the TRA Can Help in Improving the Process of Exposure Assessment Evidence that effective targeting can be implemented at Tier 1 Application of an integrated model delivers major ef?ciency,r gains across stakeholders andjurisdictions Alignment with IUCLID data requirements creates opportunities for wider harmonisation As wellasaccountingforregion speci?cvariations Widespread use ofthe TRA is generating increased stal-ceholder con?dence in its ?tness for purpose Testingforvalidation andreliabilitvremain ongoing Asimple risl-c-based tool helps identifyr where further datafscience advisable and can be used to target {leveraged} actions Enabling Effective Targeting Uses of Eonoern Demonstrateademmjrof ris koontrols or implamnt further rislr. remotion e.g. substitution Screening model identifiesthe potential 1 fromthe EL: eted hig ertier analysis: EPA or other modelse.g.Consexp-o 48ll?ugc Summary - Effective targeting releases the ef?ciencies that tiered approaches to RA potentially.?r present - Kev to targeting are workable schemes fordescribing USE Comprehensive. understandable. pragmatic, implementable Global harmonisation is an area of discussion at DEED - The TRA has demonstrated that Reliable regulatory.r decisions can be made with the minimum of data Risks can be assessed in a tiered. integrated approach and communicated in a manner that is relevant and understandable Engagement within and across industrvi'stal-teholders is essential for ensuring effective and relevant decision mal-ting ?4 What is ECETOC - WHO recognized NED funded by over 50 major companies with an interest in the safe use of chemicals - Track record ofpositive scientific engagement with EU regulators {and Hazard assessment:Classi?cation criteria: environmental fate - Not viewed bv Commissionistalteholders as having vested interests Recommendations seen to be predominantly,r based on science - Consensual approach to problem solving Transparency.r in demonstrating rationale for preferred solutions - Willingness and ability to think outside the box Avoid constraints of historical 'mistaltes' Seed the introduction of new ideas and approaches as Web Tool TM 91'! I'm! um mm lorgeie? rain Asse'E-srnem Ml? law ?HL?ml?l? raid 1 .2124?. I HA 111::th 1 inn..? D?lmlm "ru ?inhumane-51': fdht?lchil?ft . 11-1-- Ftacvl :r Tran-:91. ?nun-1' I ?i'imm am i .1. Min-Erna l: ?vinyl-.331 hi. dt?mntm r4: rcn-?c. 1.11.- 2 1r d.1 app-warn ?ux-f -i.1 'alm fo?un-? 3111? 1 I'I-ld ml?: of" ralw-i-aptml. luau-Pau- .15. hln?lndl?: H.1- apeulm p-agnuur IM- [Im- H'Ihl lurid Ann-1 lug-all?: 51M nl .. .. ..1lu' DH n: rmhuuqn: urn(unrumnwnumcrn ?mm-r4 rn amull?k?'lhul 1'43 EL 1-1-1491 uly 31.1 IN?rud ?aw $991de ??an 5r" lH'rnI-Fa-nxb'i 9:211'430 1 Jullb'o In 1 5-3-5414 ?nal-"NW no.1 .Ich-h?uhi-Ibd :HH-vu nun?angnnu BEAM 25 49|Page