Risk Management Evaluation Endosulfan

104.The “microporous” particles enable the combination of more than two active pharmaceutical ingredients into one pharmaceutical with desirable ratios to maximize the effect. The microporous particle technology also enables delivery efficiency and targeted delivery in the lungs. The manufactured pharmaceutical products currently marketed are for the treatment of patients with chronic obstructive pulmonary disease (COPD), cystic fibrosis (CF). Research on additional pharmaceutical applications is ongoing in early and late stage development (Information from IFPMA at POPRC-13). 30

105.Extensive efforts have been made to identify alternative agents, with at least 15 agents screened, but PFOB was found to be the only one suitable for manufacture of the “microporous” pharmaceutical products and to have a suitable toxicological profile that is safe for administration in humans (NDA 020-091 FDA approval of Imagent®). Given these efforts, it is improbable that an alternative agent can be identified without compromising the properties of the “microporous” particles. If an alternative agent was found, this would still require repeat clinical trials and re-registration of the products which total period will be in excess of 10 years. For this type of pharmaceutical products, there is need to secure continuous delivery to patients, hence further consideration on the appropriate way to address this application is required (Information from IFPMA at POPRC-13). 30

106.N-Ethyl perfluorooctane sulfonamide (known as sulfluramid; EtFOSA; CAS No: 4151-50-2) has been used as an active ingredient in ant baits to control leaf-cutting ants from Atta spp. and Acromyrmex spp. in many countries in South America as well as for control of red imported fire ants, and termites (UNEP/POPS/POPRC.6/13/Add.3/Rev.1). Fluorosurfactants may also be used as “inert” surfactants (enhancers used in pesticide formulations but not constituting active ingredients) in pesticide products (UNEP/POPS/POPRC.12/INF/15/Rev.1). 31

107.Sulfluramid is used in baits for the control of leaf-cutting ants, according to the delegation from Brazil the use of sulfluramid in Brazil prevents damage corresponding to losses of up to 14.5 % of trees per hectare. Other agricultural products likely to suffer costly losses are soybean and maize. Also, the per-hectare capacity to support livestock is likely to decrease if forage for grazing is reduced by ants (UNEP/POPS/POPRC.12/INF/15/Rev.1). 31

108.Insect baits for control of leaf-cutting ants from Atta spp. and Acromyrmex spp. are listed as acceptable purpose for the production and use of PFOS, its salts and PFOSF in Annex B (UNEP/POPS/POPRC.12/INF/15/Rev.1). 31

109.In a HELCOM report, cost-effective management options to reduce discharges, emissions, and losses of hazardous substances including PFOA have been assessed. Besides measures at industrial sources, measures at urban sources can also reduce emission of PFOS/PFOA, such as advanced treatment of municipal waste water by activated carbon (further details also in other options see HELCOM, 2013). 31

110.PFOA has already been phased out widely in many uses, indicating that the costs of alternatives have not inhibited the PFOA substitution. Important points to consider when evaluating the costs of alternatives for any product include the following. Alternatives with a higher initial purchase cost may actually be cheaper over the whole life span of the product when durability and other factors are taken into account. Mass-production of alternatives can significantly lower their costs. The costs of initiatives to protect health and the environment are frequently overestimated in advance and later decline rapidly after the regulation is implemented. Finally, costs of environmentally sound disposal of end-of-life products are also an important factor to take into account (Ackermann and Massey, 2006). 31

111.For the EU restriction, the substitution costs according to the EU proposal for a restriction have been estimated related to: (1) fluoropolymers import and use of PTFE mixtures; (2) textiles use in the EU; (3) textiles import in article; (4) firefighting foams; (5) paper; and (6) paints and inks. The estimation was made by the industry for the current uses (worst case scenario) and for the time period after the restriction will enter into force (more realistic case). Due to the lack of data, estimation associated with the import of PFOA in articles, photographic applications and semiconductors was not made. The estimated substitution costs for the EU range from 1.39 to 158.44 million euros with a 34.7 million euros central estimate for the more realistic case for the EU (see ECHA, 2015a, Table F.26). 31

112.The EU public consultation with industry has shown that the main fluoropolymer manufacturers have already developed several alternatives to replace PFOA. These alternatives are often exclusively manufactured and used by each company. As a consequence, there are usually no market prices available (yet). However, there are some indications on the increase in operating costs, which can be used to assess the costs of the proposed restriction to fluoropolymer manufacturers. Accordingly, it is assumed that the use of alternatives induces a low to moderate increase in production costs (0-20%). This increase arises from the higher costs and/or the higher amounts of alternatives that will be used. Industry stated that there is no change in the quality of the PTFE manufactured with the alternatives (ECHA, 2015a). 32

113.Regarding the investment costs, mainly (former) manufacturers of PFOA and PFOA-related substances industry, stated during the preparation of the EU restriction, that industry has already invested considerable resources to develop short-chain PFASs in terms of R&D efforts and capital (over 500 million euros have been reported, which was also confirmed in the EU public consultation). For downstream users, substantial costs can be expected to switch to short-chain alternatives due to reformulation of products, adapting production processes and testing. In this respect, up to 1 million euros per company have been reported, depending on the specific conditions of the case at hand (ECHA, 2015a). 32

114.According to I&P Europe, the primary barrier to completely eliminate the use of PFOA-related compounds at this time remains to be technical. However, the costs of research and development are also relevant for consideration, since such investment may represent a significant financial burden during the time when imagining industry is focused on the creation of innovative new digital imagining technologies. The economic costs associated with the substitution of PFOA-related compounds in the few remaining relevant photographic uses have in most cases become prohibitive. The small remaining relevant uses are niche products in markets that IP Europe members anticipate to further decline (I&P Europe, 2016b). 32

115.For the EU, it has been shown that there are considerable costs to society connected with hypercholesterolemia, developmental toxicity and cancer. These costs will manifest through direct costs such as medical treatment and indirect costs such as loss of life quality for affected individuals. It has not been possible to estimate the share of the overall disease burden that can be attributed to PFOA and PFOA-related substances. However, the large risk characterization ratios imply that there will be significant benefits to human health from restricting PFOA and “PFOA-related substances” (ECHA, 2015a). According to the information from Norway, the socio-economic assessment in the EU emphasized mostly on the persistent, bioaccumulative and toxic (PBT) properties of PFOA for the reason of reducing the emissions. Newer studies have also shown correlations between exposure to PFOA and reduced effects of vaccines and PFOA is presumed to be an immune hazard to humans (see e.g. UNEP/POPS/POPRC.12/11/Add.2. or NTP, 2016). 32

116.The EU restriction is not expected to lead to wider economic impacts within Europe because the market is already developing towards replacing PFOA and PFOA-related substances. This is reflected by the estimated moderate compliance costs. Furthermore, the restriction is not expected to trigger effects with regard to the competitiveness of the EU and global industry because both will have to substitute PFOA and “PFOA-related substances” to comply with the restriction. The restriction is not expected to have major effects on employment in the EU (ECHA, 2015a). 32

117.The cost of removing and destroying PFOS that is present in existing products, such as firefighting systems, is generally estimated to be well below 1,000 euros per kilogram, although it can be much higher in individual cases. An example is the Barendrecht railway tunnel in the Netherlands, where at least 3,500 euros per kilogram was spent on removing PFOS from the firefighting system. This operation did not achieve complete removal, which would require a second round of flushing or replacement of the main pipe, costing at least another 400,000 euros per kilogram of the remaining PFOS. The railway operator was not required to do this, which could be interpreted as an indication that the cost was considered disproportionate, although this reason was not put forward explicitly (Oosterhuis et al., 2017). 32

118.A regulatory initiative has been developed as part of Canada’s Chemical Management Plan (CMP) with the objective to protect the environment from risks associated with the manufacture, use, sale, offer for sale or import of (among other substances) PFOA and long-chain PFCAs. In the Canadian risk management process, scientific evidence has demonstrated that PFOA and long-chain PFCAs are persistent, that they accumulate and biomagnify in terrestrial and marine animals, and that they are toxic to the environment under the Canadian Environmental Protection Act, 1999 (CEPA). Although no quantitative analysis of benefits of the initiative has been conducted, the regulatory controls for PFOA and long-chain PFCAs in Canada will protect the environment. An improvement in environmental quality is expected from controlling these substances. 33

119.Norway states that control measures will have positive impacts on human health, since we are still exposed to PFAS in our everyday environment (Norway, 2016). The number of consumer products containing PFOA has decreased, and the levels in all-weather clothing have decreased after the introduction of a national regulation of PFOA in consumer products in 2013 (Norway Comments on 2nd draft RME). 33

120.In Australia, societal impacts of PFOA have recently come to the fore with the identification of a number of sites contaminated by the historic use of AFFFs at airports and firefighting training facilities to fight liquid fuel fires. Firefighting foams containing PFOA, PFOS and PFHxS have been phased out in a range of uses. It is noted that legacy use of AFFFs has contaminated some defense and civil airport sites, with contamination migrating off-site in some instances through surface and groundwater. The migration of PFOA from the point of use has resulted in the contamination of ground and surface water in adjoining areas that, in some instances, were used for human consumption and agricultural purposes. In sites where drinking water has been contaminated, an alternative source of drinking water has been provided. Some agricultural activities have been affected, for example, market gardens and small scale poultry and egg production, where PFOA has contaminated water previously used for these purposes. The stigma of being in a contaminated environment has led to decreasing property and business values and the loss of income for some land and business owners. This in turn has led to a level of stress and anxiety in the affected communities which is further compounded by the uncertainty of the health impacts of the residents. While the impact on Australia is largely from the legacy use of PFOA-containing AFFFs, the implementation of control measures will provide some assurance to Australian communities that the potential for ongoing or future contamination is being minimized (Australia, 2016). In April 2017, two major spills of PFOA (22,000 and 5,000 litres) containing AFFFs occurred at Brisbane airport and resulted in government warnings to avoid consuming fish from the area’s waterways (IPEN Comments on 2nd draft RME). The Australian federal government is developing a whole-of-government response and also working in collaboration with Australian States and Territories to manage and respond to PFAS contamination (Australia Comments on 2nd draft RME). 33

121.Continued use of PFOA in firefighting foams would result in the ongoing contamination of groundwater and soil surrounding military sites and airports across the world, with all its associated remediation, compensation and legal costs in addition to harms to human health and the environment (Wang et al., 2017; LaSalle, 2016; The Senate Foreign Affairs, Defence and Trade, 2016; Air Services Australia, 2016; Filipovic et al., 2015; Houtz et al., 2016). Recent calculations of the total costs for cleaning up groundwater polluted by PFAS around firefighting areas in Norway show that
3.5-5.5 million euros is required per training site. These numbers include investment and operation of groundwater cleaning systems necessary in some Norwegian airports polluted by PFAS from firefighting foams. Chemical analysis show that PFOA migrates into the ground water to a higher extent than PFOS (Norway Comments on 3rd draft RME). 33

122.In 2005 PFAS containing firefighting foam has been used at the German Airport Düsseldorf because of a plane crash and firefighting trainings. PFAS (also PFOA), contaminated soil and leached into groundwater. The PFAS containing groundwater polluted two lakes nearby which are now closed for the public, the consumption of fish is prohibited. In 2007 the local environment authority of Düsseldorf found elevated PFAS levels in the north of Düsseldorf. In the next years the airport Düsseldorf was found to be the main PFAS-source. The remediation of the groundwater will take years or even decades. Further, about 3000 tonnes of soil polluted with PFAS were excavated and disposed of. Other airports in Germany have similar PFC contaminated areas resulting from the use of AFFFs for training purposes in the past (i.e. Nürnberg airport). The costs of such remediation actions are discussed in ECHA, 2015a (Germany Comments on 3rd draft RME). 34

123.In Germany, there are is one prominent case showing the consequences of (illegal) disposal of waste/sludge on agricultural fields. Because of the disposal of industrial sludge PFOA leached into the surrounding surface water and a drinking water reservoir, Lake Möhne, was polluted (see Skutlarek et al. 2006, Wilhelm et al. 2009, Wilhelm et al. 2010, Hölzer et al. 2008, Hölzer et al. 2009). The drinking water thus contained elevated levels of PFOA. Thus, human biomonitoring studies showed higher PFOA levels in blood from people living in Arnsberg compared to inhabitants of a nearby area which received drinking water from a different source. According to information from the media the purification costs for the groundwater of about 2.5 million euros have been incurring since 2006. The purification plant will be operated during the next years and operating costs are about 100,000 euros per year (Germany Comments on 3rd draft RME). 34

124.High levels of PFAS in drinking water, in the µg/L range, have been detected since 2011 in a number of municipalities in Sweden. The costs for addressing PFAS contamination of drinking water for some municipalities are provided such as charcoal filtering of water in Uppsala (annual cost 1 million euros) and new water supply in Ronne by (3 million euros) (Swedish Environmental Protection Agency, 2016). Firefighting training sites have been shown to be the main sources of this pollution, which in some cases have resulted in water supplies being closed. The municipalities have released information that wild fish caught from lakes downstream pollution area should not be eaten too often (Swedish Chemicals Agency, 2013). For PFAS-containing water derived from a cavern near an old airfield, a carbon filter system has been installed to clean 150-200 m3 of water from the caverns before it flows out into receiving waterways (Defoort et al. 2012). PFAS have also contaminated drinking water for 15 million inhabitants and several sites in USA. However, carbon filter systems may not work for all PFAS (Wang et al., 2017). 34

125.Regarding professional, technical and protective textiles, the sales of German manufacturers in 2013 amounted up to 6 billion euros (see VTB SWT, 2016 and TM, 2016). 35

126.Due to accessibility or costs of alternative technologies, some of alternative technologies for PFOA in developing countries may be available a few years later 35

127.Due to concerns about the impact of long-chain perfluoroalkyl acids (PFAAs) on humans and the environment these PFAAs and their precursors are being substituted in many applications by other substances, including fluorinated alternatives which are structurally similar to the substances they replace. These fluorinated alternatives comprise particularly short-chain PFAAs and functionalized perfluoropolyethers (PFPEs), in particular per- and polyfluoroether carboxylic acids and (PFECAs) and per- and polyfluoroethersulfonic acids (PFESAs) having an acidic functional group attached to a per- or polyfluoroether chain instead of a perfluoroalkyl chain (Wang et al., 2015). An overview of some known fluorinated and non-fluorinated alternatives for different industry branches is given in the reference documents (ECHA, 2015a, Table C.1-1; see UNEP/POPS/POPRC.13/INF/6; Section 3 and UNEP, 2017). 35

128.The following paragraphs discuss sector specific aspects related to alternatives. Nevertheless, several aspects related in particular to risks of the alternatives (e.g. short-chain fluorinated substances) cannot be assigned to a single sector, but apply to all of those sectors where the respective alternatives are relevant. 35

129.According to FluoroCouncil, there are various alternative polymerization processing aids (PPA) used for replacing PFOA in the manufacture of fluoropolymers (FluoroCouncil, 2016a). 35

130.Fluoropolymer producers used ammonium or sodium perfluorooctanoate (APFO and NaPFO) as processing aids in the (emulsion) polymerization of polytetrafluoroethylene, perfluorinated ethylene-propylene copolymer, perfluoroalkoxy polymer and certain fluoroelastomers. In addition, ammonium perfluorononanoate (APFN) was applied in the emulsion polymerization of polyvinylidene fluoride (Prevedouros et al., 2006). Most producers have developed their own alternatives. Commercialized fluorinated alternatives are functionalized PFPEs including amongst others ADONA from 3M/Dyneon (CF3OCF2CF2CF2OCHFCF2COO-NH4+; CAS No: 958445-44-8; Gordon, 2011), GenX from DuPont or C3 Dimer salt(CF3CF2CF2OCF(CF3)COO-NH4+; CAS No: 62037-80-3; Du Pont, 2010), cyclic or polymeric functionalized PFPEs from Solvay (Marchionni et al., 2010; Pieri et al., 2011; Spada and Kent, 2011) as well as EEA-NH4from Asahi (C2F5OC2F4OCF2COO-NH4+; CAS No: 908020-52-0; EFSA, 2011a). Additional information on alternatives to PFOA in fluoropolymer production with emphasis on the manufacture of fluoropolymers in China and fluorinated emulsifier-free aqueous emulsion polymerization processes is compiled in section V of FOEN, 2017. 35

131.Three PFOA-alternatives with ether moieties (GenX,ADONA and EEA-NH4) that are generally shorter and/or less fluorinated were assessed in the EU restriction process (ECHA, 2015a, section C3). C3 Dimer salt, ADONA and EEA-NH4 are applied as alternatives for the use of PFOA as polymerization processing agent where it is applied as emulsifying agent enabling reactants from the aqueous phase and reactants from the hydrophobic phase to get into contact in an emulsion and react with each other (ECHA, 2015a). According to ECHA most of the stakeholders stated that there are no technical differences between fluoropolymers produced with the alternatives and fluoropolymers produced with PFOA (or stakeholders do not know whether there are any differences) (ECHA, 2015a). Fluoropolymer manufacturers stated during the EU public consultation that the production costs varied from none to 20% increase when applying the alternatives (ECHA, 2015a). The increase is a result of higher costs of the alternatives as well as higher amounts of the alternatives needed to manufacture one unit of fluoropolymer. Some downstream users mentioned that no cost effects occurred after substitution from PFOA to alternatives. 36

132.Toxicokinetic data of C3 Dimer salt indicate little or no metabolism, but rapid excretion. It is presumably cleared non-metabolized within 2-7 days (mouse), 10-11 h (monkey) and 4-48 h (rat). C3 Dimer salt is classified as skin irritating and eye damaging. Moreover, repeated administration resulted in liver enlargement and hepatocyte hypertrophy as well as liver cell necrosis at 0.5 mg/kg/day in male mice. With respect to carcinogenicity, a two-year rat study gave tumors at higher doses (≥50 mg/kg/day). With regards to environmental risks (data were taken from the registration dossier) related to C3 Dimer salt, it was concluded that the substance is probably not acutely toxic (LC/EC50>100 mg/L) or chronically toxic (NOEC>1 mg/L) to aquatic organisms. Regarding all available information a full PBT assessment including assessment of the criteria persistence, bioaccumulation and toxicity according to the EU chemicals legislation (for guidance see ECHA, 2017a) cannot be performed. However, the registrant acknowledges in the chemical safety report (CSR) that the C3 Dimer salt fulfils the P and the T criterion based on specific target organ toxicity after repeated exposure (STOT RE 2). The C3 Dimer salt is likely to fulfil the PBT criteria of the European chemicals legislation, see REACH Annex XIII (ECHA, 2015a). 36

133.With respect to ADONA, it turned out that the substance is persistent. No data related to carcinogenicity were available. Concerning environmental risks (data were taken from the registration dossier under the REACH regulation) related to ADONA it was concluded that the substance is probably not acutely toxic (LC/EC50>100 mg/L) or chronically toxic (NOEC>1 mg/L) to aquatic organisms. Regarding all available information a full PBT assessment cannot be performed. The substance will most probably fulfil the P criterion of REACH Annex XIII. Based on the data for environmental toxicity, the substance does not fulfil the T criterion. The registration dossier lacks toxicological information relevant to humans. Thus the data are not sufficient to conclude or to refute on the PBT-properties of the substance (ECHA, 2015a). Based on a document from the European Food Safety Authority from 2011, 3M reported that the elimination half-life of ADONA was between 12 and 34 days from the bodies of three workers, while it takes about four years in humans to clear half of the PFOA (see The Intercept, 2016 and EFSA 2011b). 36

134.In another study (Gordon, 2011) the toxicity of ADONA was evaluated in acute and
repeated-dose studies of up to 90 days, in eye and skin irritation, dermal sensitization, genotoxicity, and developmental toxicity studies. The substance was evaluated as a peroxisome
proliferator-activated receptor alpha (PPARα) agonist in rats, moderately toxic orally and practically non-toxic dermally in acute rat studies. In rabbits ADONA turned out to be a mild skin irritant and a moderate to severe eye irritant as well as a weak dermal sensitizer in local lymph node assays in mice. Based on the weight of evidence from five assays, ADONA was not considered genotoxic. No developmental toxicity was observed except at maternally toxic doses. Regarding ADONA as a PPARα agonist the liver was the primary target organ in male rats and the kidney in female rats. It was concluded by the author that the toxicity profile for ADONA is acceptable for its intended use as PPA and is superior to the one of APFO. 36