Document VJnVD0n6oQE46B2baZekxXNe8

Chemours Thermal & Specialized Solutions (TSS) contribution to the universal PFAS restriction ECHA Public Consultation 21 September 2023 Attachment 3 Table of Contents Annex I: Annex II: Annex III: Annex IV: Annex V: Complement to the information presented in Attachment 1 Chapter 1 - Go to Page Specific comments and observations on Annex A of the PFAS restriction dossier - Go to Page Comments on the analytical methods described by the Dossier Submitters in Annex E of the PFAS restriction dossier - Go to Page Socio-Economic Analysis and Impact Assessment of a potential REACH Restriction on F-gases as PFAS, report for Chemours Thermal Specialized Solutions (TSS) [attached in confidential version] - Go to Page Evaluation of Costs based on Restriction Derogations for Mobile Air Conditioning Maintenance - Go to Page Annex I: Complement to the information presented in Attachment 1 Chapter 1 Supporting information - Attachment 1, Section 1.2 Table 1 - Summary of available toxicity data for TSS substances and TFA Attribute Substance Name EC Number CAS Number Occupatio nal Exposure Limit Acute Inhalation Lethality HFO-1234yf 2,3,3,3tetrafluoropr opene 468-710-7 754-12-1 500 ppm (8hr TWA) TERA WEEL 4-hr LC50 > 405,000 ppm (> 1887 mg/L), rat 1-hr NOAEC > 100,000 ppm (> 466 mg/L), rabbit HFO1336mzz- E (2E)1,1,1,4,4,4 Hexafluoro but-2-ene 811-2130 6671186-2 400 ppm (8-hr TWA) TERA WEEL 4-hr LC50 > 17,000 ppm (> 114 mg/L), rat HFO1336mzz- Z (2Z)1,1,1,4,4,4 Hexafluoro but-2-ene 700-6517 692-49-2 500 ppm (8-hr TWA) TERA WEEL 4-hr LC50 > 102,900 ppm (> 690 mg/L), rat HFC-134a HFC-125 HFC-143a HFC-152a 1,1,1,2tetrafluoroe thane 1,1,1,2,2pentafluoroe thane 1,1,1trifluoroeth ane 1,1difluoroeth ane 212-377-0 811-97-2 206-557-8 354-33-6 206-9965 420-46-2 200-8661 75-37-6 OCCUPATIONAL EXPOSURE LIMITS 1000 ppm (8-hr TWA) - TERA WEEL 1000 ppm (8-hr TWA) TERA WEEL 1000 ppm (8-hr TWA) - TERA WEEL 1000 ppm (8-hr TWA) - TERA WEEL ACUTE TOXICITY 4-hr LC50 > 359,300 ppm (> 1500 mg/L), rat 4-hr LC50 > 800,000 ppm (> 3927 mg/L), rat 4-hr LC50 > 591,000 ppm (> 2030 mg/L), rat 4-hr LC50 > 437,500 ppm (> 1182 mg/L), rat HFC-227ea 1,1,1,2,3,3,3heptafluoropr opane 207-079-2 431-89-0 No data 4-hr LC50 > 788,696 ppm (> 5485 mg/L), rat HFC4310mee 1,1,1,2,2,3,4, 5,5,5decafluorope ntane 420-640-8 138495-428 225 ppm (8hr TWA), 700 ppm (15-min STEL) TERA WEEL 4-hr LC50 > 359,300 ppm (> 1500 mg/L), rat TFA Trifluoroacet ic Acid 200-929-3 76-05-1 4-hr LC50 > 64 ppm (> 0.3 mg/L), rat Acute Cardiac Sensitizati on Dermal Irritation Eye Irritation Skin Sensitizati on In vitro In vivo Inhalation Oral NOAEL = 120,000 ppm LOAEL > 120,000 ppm None None None Negative (mutagenicit y & clastogenicit y) Negative (clastogenici ty) 13-week NOAEC = 50,000 ppm, rat* 4-week NOAEC = 10,000 ppm, minipig* 4-week NOAEC = 500 ppm, rabbit Not a relevant route of exposure. NOAEL = 70,000 ppm LOAEL > 70,000 ppm None None None Negative (mutageni city & clastogeni city) Negative (clastogeni city) 13-week NOAEC = 10,000 ppm, rat Not a relevant route of exposure. NOAEL = 12,500 ppm LOAEL = 25,000 ppm None None None Negative (mutageni city & clastogeni city) Negative (clastogeni city) 13-week NOAEC = 5000 ppm, rat Not a relevant route of exposure. NOAEL = 50,000 ppm LOAEL = 75,000 ppm None NOAEL = 75,000 ppm LOAEL = 100,000 ppm None NOAEL = 250,000 ppm LOAEL = 300,000 ppm None None None None None None None GENOTOXICITY Negative (mutagenici ty & clastogenici ty) Negative (clastogeni city) Negative (mutagenicit y & clastogenicit y) Negative (clastogenici ty) Negative (mutagenic ity & clastogenic ity) Negative (clastogeni city) REPEAT DOSE TOXICITY 13-week NOAEC = 50,000 ppm, rat* 13-week NOAEC = 50,000 ppm, rat* 13-week NOAEC = 40,000 ppm, rat* Not a relevant route of exposure. Not a relevant route of exposure. Not a relevant route of exposure. NOAEL = 50,000 ppm LOAEL = 150,000 ppm None None None Negative (mutagenic ity & clastogenic ity) Negative (clastogeni city) 2-year NOAEC = 25,000 ppm, rat* Not a relevant route of exposure. NOAEL = 70,000 ppm LOAEL = 90,000 ppm None None None Negative (mutagenicity & clastogenicity ) Negative (clastogenicit y) 13-week NOAEC = 105,000 ppm, rat* Not a relevant route of exposure. NOAEL = 50,000 ppm LOAEL = 75,000 ppm None None None Negative (mutagenicit y & clastogenicit y) Negative (clastogenicit y) 13-week NOAEC = 1500 ppm, rat** Not a relevant route of exposure. no data Corrosive Corrosive None (readacross) Negative (mutagenicit y & clastogenicit y) no data Not a relevant route of exposure. 13-week NOAEL = 8.4 mg/kg, rat 52-week NOAEL = 37.8 mg/kg, rat* Developm ental Toxicity Reproduc tive Toxicity Not a developme ntal toxicity hazard, rat & rabbit Fetal NOAEC = 50,000 ppm (rat developmen tal)* Fetal NOAEC = 4000 ppm (rabbit developmen tal) Not a reproductiv e toxicity hazard, rat NOAEC = 50,000 ppm (twogeneration repro)* Not a developm ental toxicity hazard, rat Fetal NOAEC = 7500 ppm Not a reproduct ive toxicity hazard, rat NOAEC = 12,500 ppm (twogeneration repro)* DEVELOPMENTAL & REPRODUCTIVE TOXICITY Not a developm ental toxicity hazard, rat & rabbit Fetal NOAEC = 1500 ppm (rat developm ental) Fetal NOAEC = 7500 ppm (rabbit developm ental) Not a reproduct ive toxicity hazard, rat NOAEC = 2500 ppm (twogeneration repro)* Not a developme ntal toxicity hazard, rat & rabbit Fetal NOAEC = 50,000 ppm (rat developme ntal)* Fetal NOAEC = 40,000 ppm (rabbit developme ntal)* Not a reproducti ve toxicity hazard, rat & mouse NOAEC = 50,000 ppm (rat twogeneration repro)* NOAEC = 50,000 ppm (male mouse dominant lethal)* Not a developme ntal toxicity hazard, rat & rabbit Fetal NOAEC = 50,000 ppm (rat developmen tal)* Fetal NOAEC = 50,000 ppm (rabbit developmen tal)* Not a reproductive toxicity hazard, rat (readacross) Not a developm ental toxicity hazard, rat & rabbit Fetal NOAEC = 40,000 ppm (rat developme ntal)* Fetal NOAEC = 40,000 ppm (rabbit developme ntal)* Not a reproductiv e toxicity hazard, rat (readacross) Not a developm ental toxicity hazard, rat Fetal NOAEC = 50,000 ppm (rat developme ntal)* Not a reproductiv e toxicity hazard, rat (readacross) Not a developmen tal toxicity hazard, rat & rabbit Fetal NOAEC = 105,000 ppm (rat development al)* Fetal NOAEC = 105,000 ppm (rabbit development al)* Not a reproductive toxicity hazard, rat (read-across) Not a developmen tal toxicity hazard, rat Fetal NOAEC = 3500 ppm (rat development al)* Not a reproductiv e toxicity hazard, rat NOAEC = 3500 ppm (onegeneration repro)* Not a developme ntal toxicity hazard, rat Fetal NOAEL = 150 mg/kg (rat developmen tal)* Developme ntal toxicity hazard, rabbit Fetal NOAEL not established, LOAEL = 180 mg/kg (rabbit developmen tal)*** Not a reproductiv e toxicity hazard, rat NOAEL = 242-265 mg/kg (extended onegeneration repro)* Acute Aquatic Toxicity Chronic Aquatic Toxicity Biotic Degradati on Soil Adsorptio n Coefficien t (LogKoc) 96-hr LC50 > 197 mg/L, fish 48-hr EC50 > 100 mg/L, daphnia 72-hr EC50 > 100 mg/L, algae No data Not readily biodegradab le no data 96-hr LC50 = 1.7814.1 mg/L, fish 48-hr EC50 = 92.9 mg/L, daphnia 72-hr EC50 > 14.4 mg/L, algae NOAEC = 0.131 mg/L, fish NOAEC = 0.186 mg/L, daphnia NOAEC = 4.24 mg/L, algae Not readily biodegrad able -0.27-0.47 96-hr LC50 76.1 - > 95.7 mg/L, fish 48-hr EC50 = 22.5 mg/L, daphnia 72-hr EC50 > 23.7 mg/L, algae NOAEC = 9.59 mg/L, fish NOAEC = 10.2 mg/L, daphnia NOAEC = 6.92 mg/L, algae Not readily biodegrad able 2.51 ECOTOXICITY 96-hr LC50 = 450 mg/L, fish 48-hr EC50 = 980 mg/L, daphnia 72-hr EC50 > 100 mg/L, algae (readacross) 96-hr LC50 > 100 mg/L, fish (readacross) 48-hr EC50 > 100 mg/L, daphnia (read- across) 72-hr EC50 > 100 mg/L, algae (read- across) 96-hr LC50 = 40 mg/L, fish 48-hr EC50 = 300 mg/L, daphnia (readacross) 72-hr EC50 > 100 mg/L, algae (readacross) No data No data No data ENVIRONMENTAL FATE Not readily biodegrada ble Not readily biodegradab le Not readily biodegrada ble 1.57 (estimated) 1.30 (estimated) no data 96-hr LC50 > 100 mg/L, fish (readacross) 48-hr EC50 > 100 mg/L, daphnia (readacross) 72-hr EC50 > 100 mg/L, algae (readacross) No data Not readily biodegrada ble 0.7 96-hr LC50 > 100 mg/L, fish (readacross) 48-hr EC50 > 100 mg/L, daphnia (read-across) 72-hr EC50 > 100 mg/L, algae (readacross) No data Not readily biodegradabl e 2.3 96-hr LC50 = 450 mg/L, fish 48-hr EC50 = 980 mg/L, daphnia 72-hr EC50 > 100 mg/L, algae (readacross) No data Not readily biodegradabl e 2.3 96-hr LC50 = 999 mg/L, fish 48-hr EC50 = 999 mg/L, daphnia 72-hr EC50 = 237 mg/L, algae (readacross) NOAEC = 3.8 mg/L, fish NOAEC = 25 mg/L, daphnia NOAEC = 5.6 mg/L, algae Not readily biodegradab le 0.8 *highest exposure/dose tested ** Based on an assessment of all relevant inhalation data, the health effect of highest concern is a transient CNS effect with a NOAEL of 1500 ppm and a LOAEL of approximately 2000 ppm. *** According to the dossier submitters, the findings seem to be rabbit specific since comparable findings were not observed in a developmental toxicity study in rats at comparable dose levels (Baxter, 2010; Chunsheng, 2020). Further investigations are underway to elucidate the reason for the species differences and potential essential mechanism of action to fully understand the relevance of the findings for human health. The submitter participates in a partnership with other companies ("TFA Task Force") which operate under the EU Plant Protection Product Framework to collectively contribute to the reprotoxicological assessment of TFA. A conclusion on the classification and labelling of TFA will be postponed until the final results of the studies investigating the effects on development are available. Additional considerations on the available toxicity data for HFO-1234yf, HFO-1336mzzZ and TFA HFO-1234yf. HFO-1234yf is not classified for any human or environmental health endpoint according to CLP Regulation and its amendments. HFO1234yf is not considered an acute inhalation hazard as is demonstrated by rat and mouse 4-hour inhalation lethal concentrations (LC50s) of greater than 405,000 and 99,830 ppm, respectively and a 1-hour LC50 in rabbits of greater than 100,000 ppm. No cardiac sensitization was observed at the highest concentration tested (120,000 ppm) in dogs. Irritation and sensitization studies are not technically feasible because the substance is a gas at room temperature and ambient pressure. However, no signs of eye, skin or respiratory irritation or sensitization were observed in several inhalation toxicity studies. These include rabbits exposed by whole-body for up to an hour at 100,000 ppm (466,000 mg/m3), rats exposed noseonly for 4 hours at exposure concentrations up to 405,000 (1,887,300 mg/m3), and rats (nose-only) and rabbits (whole-body) repeatedly exposed to concentrations as high as 50,000 ppm (233,000 mg/m3) for four or thirteen consecutive weeks. No toxicity was observed in male or female rats exposed by inhalation to concentrations up to 50,000 ppm for 6 hours a day for 2, 4, or 13 weeks. In a 28-day inhalation toxicity study in rabbits, subacute/chronic cardiac inflammation was observed in males exposed to greater than 500 ppm (equivalent to 2330 mg/m3) and females exposed to greater than 1000 ppm (equivalent to 4660 mg/m3). Exposure was for 6 hours a day, 7 days a week. However, as the rabbit is not an appropriate model for human cardiotoxicity, a 28-day inhalation toxicity study in the minipig was conducted. No test substance-related effects were observed during the course of this study, in which animals were exposed to test concentrations up to 10,200 ppm for 6 hours a day, 7 days a week. Based on the weight of evidence, HFO-1234yf is not considered to be genotoxic in mammalian systems. Based on the lack of mutagenic activity in mammalian systems, the absence of systemic and organ toxicity, and its physical-chemical and metabolism properties, HFO-1234yf is not expected to be carcinogenic in humans. HFO-1234yf was not uniquely toxic to the developing fetus in both rat and rabbit studies and was not associated with adverse reproductive effects at concentrations up to and including 50,000 ppm in a two-generation reproduction toxicity study in rats. Based on the presented data, no quantitative conclusions can be drawn with respect to absorption, distribution, metabolism, and excretion, as no mass balance was determined in any of the in vivo studies. However, since less than 0.1% of the administered dose was recovered from urine in the form of metabolites, biotransformation appears to be quite low and most of the absorbed HFO-1234yf will have most likely been exhaled as parent compound. The dermal route of exposure is not relevant for this substance. As HFO-1234yf is a gas, it is expected to partition primarily to air. Direct and indirect exposures to water are unlikely during its life cycle. The substance is not considered to be toxic to or bioaccumulate in aquatic or air-breathing organisms. HFO-1234yf is not readily biodegradable. The substance has low toxicity to freshwater fish, invertebrates, and algae with a 96-hr LC50 greater than 197 mg/L, a 48-hr EC50 greater than 100 mg/L, and a 72-hr EC50 greater than 100 mg/L, respectively. HFO-1336mzzZ. HFO-1336mzzZ is not classified for any human health or environmental endpoint according to CLP Regulation and its amendments. HFO-1336mzzZ has low acute mammalian toxicity potential as demonstrated by a rat inhalation lethal concentration (LC50) of greater than 102,900 ppm. No cardiac sensitization was noted in dogs at an exposure concentration of 12,500 ppm. Irritation and sensitization studies are not technically feasible because the substance is a volatile liquid at room temperature and ambient pressure. Irritation studies are technically not feasible because the substance volatilizes at biological temperatures and ambient pressure. However, no signs of eye, skin, or respiratory irritation or sensitization were observed in several inhalation toxicity studies. These studies include nose-only exposures for 4 hours in rats up to 102,900 ppm, whole-body exposures for 2 hours in rats up to 21,000 ppm, whole-body repeat exposures in rats and rabbits to concentrations as high as 10,000 ppm and 15,000 ppm, respectively. HFO-1336mzzZ is consistent with a substance with very low repeated dose toxicity as demonstrated by the lack of any substance-related adverse effects at inhalation exposure concentrations up to and including 5000 ppm for 90 days. HFO-1336mzzZ exhibited no genotoxicity when tested in bacterial and mammalian in vitro systems or in rats. Based on the lack of mutagenic activity in mammalian systems, the absence of systemic and organ toxicity, and its physical-chemical and metabolism properties, HFO-1336mzzZ is not expected to be carcinogenic in humans. HFO-1336mzzZ was not uniquely toxic to the developing fetus in both rat and rabbit studies. The NOAEC for both maternal and fetal effects in rats was 1500 ppm and the NOAECs for maternal and fetal effects in rabbits were 5000 ppm and 7500 ppm, respectively. No reproductive toxicity effects were noted in a two-generation reproductive inhalation toxicity study in rats up to 2500 ppm, the highest exposure concentration. In a toxicokinetics study conducted according to OECD test guideline 417, male and female rats were exposed to the substance via inhalation and toxicokinetics parameters including absorption, distribution, metabolism, and excretion were evaluated. No bioaccumulation potential was identified in this study. As HFO-1336mzzZ is a volatile liquid, it is expected to partition primarily to air. Direct and indirect exposures to water are unlikely during its life cycle. The substance is not considered to be toxic to or bioaccumulate in aquatic or air-breathing organisms. HFO-1336mzzZ is not readily biodegradable. The substance has low acute toxicity to freshwater fish, invertebrates, and algae with a 96-hr LC50 greater than 95.7 mg/L, a 48-hr EC50 equal to 22.5 mg/L, and a 72-hr EC50 greater than 23.7 mg/L, respectively. The substance has low chronic toxicity to freshwater fish, invertebrates, and algae with chronic NOAECs of 9.59, 10.2, and 6.92 mg/L, respectively. TFA. According to CLP, Annex VI Table 3.1, TFA causes severe skin burns and eye damage, is harmful if inhaled and is harmful to aquatic life with long lasting effects. TFA is not classified for skin sensitization, specific target organ toxicity following single or repeated exposure, germ cell mutagenicity, reproductive toxicity, effects on or via lactation, or acute aquatic toxicity. TFA is not readily biodegraded in water and no biodegradation or cometabolic transformation by any of the microorganisms tested was observed under aerobic conditions. The reported log Kow value for TFA is 0.5 indicating negligible potential for bioaccumulation. It should be noted that this is not relevant at environmental pH range, where it is well-known that ionized substances are less lipophilic and less adsorbed. This is confirmed by the lower experimental value (-4.1), and the data obtained with SPARC model (-0.58 at pH 5.5 and above). Nevertheless, the theoretical log Kow of 0.79 was retained as a worst-case key value for the Chemical Safety Assessment in the REACH dossier. Studies on aquatic and terrestrial organisms indicate a low level of incorporation and/or accumulation of TFA by natural microbial communities, aquatic organisms spanning a range of trophic levels, and most terrestrial plants. With respect to mammalian toxicity, due to its acidic properties, TFA will induce local effects by all routes of exposure before systemic effects are expected. Corrosion to skin and eyes are expected and respiratory irritation has been observed in animal studies. Local irritating effects were observed in the nasal epithelium of rats after a 4-hour nose-only inhalation exposure at 300 mg/m3 TFA, however these effects were reversible after a 14-day observation period. No mortality was observed at this concentration. Because of its acidity, TFA is classified for both skin and eye corrosion/irritation and for acute inhalation toxicity according to CLP and GHS. Considering the strong corrosive properties of TFA, repeated dose toxicity testing was conducted with the neutralized form of TFA (i.e., sodium trifluoroacetate; NaTFA). To characterize the repeated dose oral toxicity of TFA, results of 28- and 90-day dietary toxicity studies and a 1-year drinking water toxicity study are available. In a GLP-compliant, 90-day repeated dose oral toxicity study conducted according to OECD test guideline 408, adverse effects were observed in the liver (increase in liver weight, histopathological changes in the liver, and changes in hematological parameters, clinical biochemistry, and urinalysis). The no observed adverse effect levels (NOAELs) from this study were determined to be 8.4 and 10.1 mg TFA/kg body weight per day in male and female rats, respectively. No treatment-related adverse effects were observed in a separate GLPcompliant drinking water study that was conducted in accordance with OECD test guideline 452. In that study, the NOAELs for male and female rats administered NaTFA via drinking water for 52 weeks were 37.8 and 64.0 mg TFA/kg body weight per day, respectively, the highest administered doses. Based on the body of toxicological evidence after repeated oral administration of TFA, the liver can be considered as the target organ. The target organ effects observed in the animal studies, which were generally mild, are consistent with a peroxisome proliferated-activated receptor alpha (PPAR alpha) model of action. In the context of the human relevance framework, PPAR alpha interaction with subsequent proliferative hepatic effects is a mode of action not considered relevant to humans1 as human liver, in contrast to rodents, does not respond to PPAR alpha activation by initiating cell proliferation. Two GLP-compliant oral prenatal developmental toxicity studies were conducted according to OECD test guideline 414, one in rats and the other in rabbits. No test substance-related adverse effects were noted in the rat study after exposure to TFA via oral gavage; the NOAEL for maternal and embryo-fetal effects was established at 150 mg/kg/day, the highest dose tested. In the rabbit study, maternal liver pathology (moderate bile duct hyperplasia/fibrosis) was noted at 375 and 750 mg NaTFA/kg/day; the maternal NOAEL was 180 mg NaTFA/kg/day. In contrast to the rat study, fetal abnormalities predominantly affecting the eyes were observed at all dose levels (LOAEL = 180 mg NaTFA/kg/day) and a NOAEL for embryo-fetal developmental toxicity could not be established in this study. Since comparable findings at comparable dose levels were not observed in the rat study, the findings appear to be rabbit specific. Therefore, further investigations are currently underway to elucidate the reason for the species differences and potential mechanism, which is essential to fully understanding the human-relevance of the findings. In a GLP-compliant extended one-generation reproductive toxicity study performed according to OECD test guideline 443 in rats, no adverse effects on reproductive 1 Felter, S.P. et al. (2018) `Human relevance of rodent liver tumors: Key insights from a Toxicology Forum Workshop on nongenotoxic modes of action', Regulatory Toxicology and Pharmacology, 92, pp. 1-7. https://doi.org/10.1016/j.yrtph.2017.11.003. performance, offspring development, or general systemic toxicity were observed after dietary administration of NaTFA. The NOAEL for this study was approximately 242-265 mg TFA/kg/day. NaTFA was not found to be mutagenic or clastogenic in a series of in vitro genetic toxicity studies. Supporting information - Attachment 1, Section 1.3 It is not correct to assume that the TFA yield from trifluoroacetaldehyde (CF3CHO) is 10%. The TFA yield from trifluoroacetaldehyde is estimated at 2%, with an upper theoretical limit of approximately 30%. In Annex B, page 50, section B.4.1.3.2, the Dossier Submitters discuss the atmospheric degradation of trifluoroacetaldehyde (CF3CHO), which can occur via three competing reactions and add that "how important the three different degradation processes are relative to each other is unclear, while up to 10% formation of TFA from trifluoroacetaldehyde (CF3CHO) has been estimated by UBA (2021) page 109."2 The 2022 EEAP ReportError! Bookmark not defined. concludes that "the TFA yield from processing of CF3CHO is estimated at 2% with an upper theoretical limit of ~ 30%." As discussed in the report, CF3CHO sinks include photolysis, reaction with OH radical and reaction with water. Photolysis of CF3CHO does not contribute to the formation of TFA and the EEAP report states that "the importance of formation of TFA from the reaction of OH with CF3CHO was indirectly accessed by Sulbk Andersen et al.3 in a global modeling study of HCFO-1233(zd). This model, which did not include potential CF3CHO-hydrate formation, suggested a 2% yield of TFA from CF3CHO". The contribution to TFA from reaction with water is highly uncertain, and the EEAP concludes stating that "the contribution to TFA from CF3CHO hydrate formation and processing remains highly uncertain (Franco et al., 2021; Sulbk Andersen et al., 2006). Assuming that uptake into cloud water and hydration is efficient, effectively converting CF3CHO into TFA on a timescale of 5 days (only limited by transport limitations, i.e., the lower limit for the time taken for transport into clouds (Wallington et al., 1994), then a maximum TFA yield of 27% can be expected from the hydrate formation". An upper theoretical limit of TFA from CF3CHO is thus 2% from OH reaction plus approximately 27% from reaction with water, leading to an upper theoretical limit of approximately 30%. There is no scientific evidence on the formation of up to 11% of HFC-23 due to photolysis of trifluoroacetaldehyde. In Annex B, page 50, section B.4.1.3.2, the Dossier Submitters cite a preprint publication by Campbell et al.4, stating that photolysis of CF3CHO might lead to the formation of up to 11% of HFC-23 (CF3H). 2 Behringer, D., et. al., (2021), Persistent degradation products of halogenated refrigerants and blowing agents in the environment: type, environmental concentrations, and fate with particular regard to new halogenated substitutes with low global warming potential, Final Report, UBA 2021, https://www.umweltbundesamt.de/sites/default/files/medien/5750/publikationen/2021-05-06_texte_732021_persistent_degradation_products.pdf 3 Sulbaek Andersen, M.P. et al. (2018) `A three-dimensional model of the atmospheric chemistry of E and ZCF3CH=chcl (HCFO-1233(ZD) (E/Z))', Atmospheric Environment, 179, pp. 250-259. https://doi.org/10.1016/j.atmosenv.2018.02.018. 4 Campbell, J., Hansen, C. and Kable, S. (2021), Photodissociation of CF3CHO provides a new source of CHF3(HFC23) in the atmosphere: implications for new refrigerants, Research Square February 2021, https://www.researchsquare.com/article/rs-199769/v1 CF3CHO is indicated as an intermediate in the tropospheric degradation of molecules containing the CF3CH= moiety. The major atmospheric fate of the intermediate CF3CHO is photolysis5. The photolysis of CF3CHO has three principal pathways6: CF3CHO + h CF3 + HCO (1a) CF3CHO + h CF3H + CO (1b) CF3CHO + h CF3 + CO + H (1c) Sulbaek-Andersen and Nielsen7 examined the photolysis of CF3CHO under tropospheric conditions and found no detectable (< 0.3 %) CF3H produced during the tropospheric photolysis of CF3CHO, concluding that "pathway 1b is of no significant importance for the photolysis of CF3CHO from 400 nm down to 290- 300 nm." In addition, their experiments indicated that most, if not all of the photolytic degradation of CF3CHO proceeds via pathway 1a. The tropospheric photolysis of CF3CHO thus occurs via pathways 1a (major) or 1c (minor) to produce the CF3 radical and the HCO radical (pathway 1a) or the CF3 radical, CO and H (pathway 1c). The atmospheric fates of HCO and CO lead to the ultimate formation of CO2. The fate of the CF3 radical leads to the ultimate formation of HF and CO2. In an earlier study, Chiappero et. al.5 examined the photolysis of CF3CHO under tropospheric conditions, and as reported by Sulbaek-Andersen and Nielsen, found that photolysis produces CF3 and HCO radicals: CF3CHO + hv = CF3 + HCO Chiappero et. al. found no evidence of CF3H formation in the photolysis of CF3CHO, in agreement with Sulbaek-Andersen and Nielsen. The results of Sulbaek-Andersen and Nielsen and of Chiappero et al. indicate that the ultimate main products of the tropospheric photolysis of CF3CHO are HF and CO2. These studies dispute the conclusion of an earlier non peer-reviewed report by Campbell, et. al.40, claiming an 11% yield of CF3H in the photolysis of CF3CHO. However, the paper was submitted to Research Square in February of 2021 and to date has not been published in a peer-reviewed journal. In their experiments, Campbell et. al. did not detect CF3H but inferred its formation based on the detection of CO. Interestingly, in a follow-up study which appeared in a peer-reviewed journal, Campbell et. al.8 examined the photodissociation dynamics of CF3CHO and did not report the formation of CF3H from the photolysis of CF3CHO; CF3CHO was reported to be photolyzed to produce the products CF3 and HCO. 5 Chiappero, M.S., Malanca, F.E., Argello, G.A., Wooldridge, S.T., Hurley, M.D., Ball J.C., Wallington, T.J., Waterland, R.L. and Buck R.C. (2006), Atmospheric chemistry of perfluoroaldehydes (CxF2x+1CHO) and fluorotelomer aldehydes (CxF2x+1CH2CHO): quantification of the important role of photolysis, J Phys Chem A. 2006 Nov 2;110(43), 11944-53, https://doi.org/10.1021/jp064262k. 6 Calvert, J., Mellouki, A., Orlando, J., Pilling, M. and Wallington, T. (2011), The Mechanisms of Atmospheric Oxidation of the Oxygenates, Oxford University Press, 2011, https://www.researchgate.net/publication/346818870_Mechanisms_of_Atmospheric_Oxidation_of_the_Oxygena tes 7 Sulbaek-Andersen, M. and Nielsen, O. (2022), Tropospheric photolysis of CF3CHO, Atmospheric Environment 272 (2022) 118935, https://reader.elsevier.com/reader/sd/pii/S1352231021007573?token=43C4D7D44D9854635E4F502A5AA266CDF 16857AC14F1FA0D0BCB22889D261C0D6B8F360FA22CBECBC726BC2BCB5609CB&originRegion=eu-west1&originCreation=20230504174017 8 Campbell, J., Nauta, K., Kable, S.H. and Hansen, C.S. (2021), Photodissociation dynamics of CF3CHO: C-C bond cleavage, J. Chem. Phys. 155, 204303, https://pubs.aip.org/aip/jcp/article/155/20/204303/280584/Photodissociation-dynamics-of-CF3CHO-C-C-bond Annex II: Chemours' TSS comments to Annex A: "Manufacture and uses" to the ANNEX XV RESTRICTION REPORT A.2. Manufacture, import and export A.2.1.7.1. PFAAs and PFAA precursors Annex A claim page 15: Table A.5. Annual volumes of fluorinated gases in PFAS scope manufactured in the EEA. Annex A claim page 17: Table A.7. PFASs manufacturing volumes in EEA (in 2020) Chemours' Response: The manufacturing of fluorinated gases in the European Economic Area is overestimated in Table A.5. and Table A.7. as shown above. The use of REACH registration midpoint volumes to estimate production and import is less robust than the official F-Gas reporting and is misleading: it is accounting for both manufacturing and import registrations. Registered tonnage band do not represent actual production or imports and is not reflecting real market size. Confronting mandatory EEA report official data (15 000 tons/yr9) with the result of 176 548 tons/yr mentioned in Table A.5 and A.7 in Annex A would mean c.a. 160 000 tons/yr have not been reported by manufacturing companies in the EU.10 A.3. Uses A.3.1 Summary Annex A claim page 21: Table A.10. Estimated tonnages for PFAS manufacture and major PFAS use sectors for 2020. Tonnages are for new products on the market, unless stated otherwise. Chemours' Response: The estimated stock of PFAS fluorinated gases in 2018, as reported in Table A.10 of Annex A, is higher than a calculation based on EEA GHG inventory 2018 (submitted 2020 v2; Table2(II)B-Hs2)11 and EEA fluorinated gas report 2020 (Table A5.17).12 In Annex A, Table A.10, non-PFAS fluorinated gases should be deducted and HFOs should be added. As stated in figure A.19 (page 78), more than 25% and 83 155 tons of the total supply of f-gases are not considered as PFAS. Relating to table A.10, it should be clearly stated 9 https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2021/fluorinated-greenhouse-gases2021-annex/view 10 Registered substances. ECHA. (n.d.). https://echa.europa.eu/information-on-chemicals/registered-substances EEA Fluorinated greenhouse gases report 2021_annex Table 1 EU production of fluorinated gases (tons) 11 CRF tables EU-27+UK. European Environment Agency. (2020, May 28). https://www.eea.europa.eu/publications/european-union-greenhouse-gas-inventory-2020/eu-27-crftables.zip/view 12 Fluorinated greenhouse gases 2020. European Environment Agency. (2020, Dec 1) https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2020 when using data that contains substances that are not considered as PFAS, or the data should be adjusted accordingly. A 3.9 Applications of fluorinated gases A.3.9.1. Uses Annex A claim page 63: Ambition of the F-gas regulation: "Limiting the use of some important F-gases that can be produced and imported into the EU; Banning the use of F-gases in many new types of equipment where less harmful alternatives are widely available; Preventing emissions of F-gases from existing equipment." Chemours' Response: Detailed information can be found in Attachment 1 Section 1.5. A.3.9.1.1. Refrigeration, air conditioning and heat pumps Annex A claim page 64: Main sub-use categories assessed: o Domestic refrigeration o Commercial refrigeration o Industrial refrigeration o Transport refrigeration o Mobile air conditioning (MAC) o Stationary air conditioning (AC) and heat pumps (HP) o Domestic air conditioning and domestic heat pumps for space heating o Commercial air conditioning and heat pumps o Domestic heat pumps (clothes dryers) Chemours' Response: Space cooling should be considered on top of space heating. Additionally, Chemours acknowledges that there is a gap in the sub-use categories listed here. There is no reference to commercial heat pumps other than air conditioning. It is important to note that heat pumps for industrial processes such as food, pharma, and petrochemicals, differ from industrial refrigeration. Annex A claim page 64: "However, often fluorine-free alternatives are available, like the natural refrigerants carbon dioxide (CO2), hydrocarbons and ammonia." Chemours' Response: It should be noted that equipment based on these alternatives is not available across the list of applications mentioned above. More specifically, Hydrocarbons based residential Heat Pumps (the only alternative to f-gases in this application segment) are submitted to use limitation by international building codes. Ammonia - due to the high toxicity- can practically only be employed in industrial applications under specific constraints in terms of safety monitoring, space access and distance from residential buildings. Transcritical CO2 equipment is not commercially available/viable in a range of medium and small refrigeration capacity applications. More detail is available in the Attachment 2, Stationary HVACR application submission. Annex A claim page 64: "In larger commercial systems CO2 is technically feasible." Chemours' Response: The downsize of CO2 transcritical equipment remains an issue, not allowing to bridge the capacity provided by large CO2 transcritical systems, down to R-290 Hydrocarbons low refrigeration capacity systems (systems availably, projected costs if developed in the future, lack of energy efficiency compared to PFAS-based systems). For example, in the range of applications below the 40kW refrigeration threshold, CO2 transcritical equipment are either unavailable or non-competitive (both in terms of Capex and energy efficiency), as down-scaling still represents a major technological challenge. More detail is available in the Attachment 2, Stationary HVACR application submission. Annex A claim page 65: "The refrigerant circuits for electric vehicles are more complex and larger (single thermal management system), as the battery must also be cooled" Chemours' Response: The above claim does not currently include temperature management needs for various electronics and power train components, on top of batteries. Furthermore, the need is not only to cool batteries, but also to heat them when ambient temperature is lower than the operational threshold. CO2 as a weak link in the overall thermal management system (due to limited cooling power at high ambient temperatures) significantly increases the risk of a "thermal runaway event"13 due to the "single thermal management" approach required to extend driving range. This is seen in the study of the heat pump for a passenger electric vehicle based on refrigerant R-744.14,15 Annex A claim page 65: "One stakeholder has pointed out that future cars and vehicles will be electrically driven and that combined air-conditioning and heat pump systems will be the standard solution due to energy efficiency constraints" Chemours' Response: The energy consumption of the heating system for electric vehicles (EV) decreases the maximum mileage. Therefore, the energy saving technology for heating system is critical for EV. The initial heating system for the EV used AC systems complemented by positive temperature coefficient (PTC) heater PTC. The PTC heater is a convenient heating method used in EV, but PTC heaters have some defects such as low efficiency. The heat pump (HP) system is gradually replacing PTC, however, HPs have other challenges for the heating capacity and efficiency in low temperature environments. 13 Energies | Free Full-Text | Review of Thermal Management Technology for Electric Vehicles (mdpi.com) 14 Canteros and Polansky, Archives of Thermodynamics, Vol 43 (2022) 15 Bosch Mobility 2023 According to the American Association of Automotive Manufacturers (SAE).16 The energy consumption of the air conditioning and the PTC material for heating accounts for 33% of the energy consumption of the whole vehicle. If electric vehicles use heat pump air conditioning systems instead of heating PTC materials to meet heating needs in winter, they will significantly increase driving range and promote the rapid development of electric vehicles. Single source heat pump AC systems are a convenient replacement and still dominant in EVs, especially in mild climate areas, with low cost and easy maintenance. However, the heat pump AC system, which only involves the necessary modifications based on conventional vehicle AC systems, has low system efficiency. More information is available in the Attachment 2, MAC and heat pump application submission. A.3.9.1.2. Fluorinated gases used as Foam-Blowing Agents Annex A claim page 66: Chemours' Response in Support of F-gas uses in Polyurethane spray foam: Additional applications of SPF are also found in flooring, marine applications, aerospace, industrial bulk tank insulation, wine cellars, agricultural livestock farms, commercial barge hull insulation, monolithic domes, sports arenas, cryogenic, vaccine transportation container insulations, and LNG tankers. The advantage of using SPF with HFOs vs. alternatives (fiberglass, Mineral Wool, water blown foams, or CO2 blown foams): SPF expands and fills cracks and seams to reduce air infiltration. SPF fully adheres to substrates and conforms to irregular shapes and surfaces. Controls indoor air quality by keeping out air pollutants. SPF prevents mold and mildew by controlling moisture. Reduce pollutants, allergens & pests entering buildings.17 They are dimensionally stable and do not sag or settle over time which maintains their insulation value for the life of a structure. They have exceptional physical properties adding to the overall strength of the structure. They are not friable or brittle so they will remain intact and in place even in harsh environmental conditions. They are water resistant and are considered a Class V flood resistant insulation. Even if the insulation gets wet, it can be dried and remain in service. More detail is available in the Attachment 2, Foam blowing agent application submission. A.3.9.1.3. Solvents Annex A claim page 67 - 68: "Particularly relevant are fluorinated solvents used for cleaning of parts and component in oxygenenriched environments. An industrial use of fluorinated solvents is the use as carrier fluids to deposit lubricants, silicones, coatings, adhesives and other materials in smooth coatings, as well as for the 16 Yokoyama, A., Osaka, T., Imanishi, Y., and sekiya, S., "Thermal Management System for Electric Vehicles," SAE Int. J. Mater. Manuf. 4(1):1277-1285, 2011, https://doi.org/10.4271/2011-01-1336. 17 SPFA-154 011021.pdf (sprayfoam.org) formulation of dissolved polymeric PFAS oils and greases. Furthermore, fluorinated gases/solvents may be used as heat transfer media, thermal testing fluids and in electrical/electronics testing. Fluorinated solvents (e.g. hexafluoroisopropanol, HFIP) are used in additive (3D) printing as a debinding agent prior to sintering for 3D printing of metals. They are also used as a smoothing agent for some polymer 3D printing applications, including for respiratory medical articles, Covid-19 diagnostic items, automotive and aerospace components, electronics and consumer items." Chemours' Response: While key applications for Fluorinated Solvents include industrial metal cleaning to remove oil and grease, electronics cleaning for the removal of flux, precision cleaning to remove particulates or dust, cleaning in relation to various lubrication processes and cleaning of parts and component in oxygen-enriched environments, these markets are highly fragmented and extremely specialized. Fluorinated solvents are only used where required and where the highest and most stringent standards are necessary. Another key application where fluorinated solvents are used is in chillers for heat transfer applications in the semiconductor manufacturing process. It is mentioned that non-flammability, thermal and chemical stability, dielectric properties (low electrical conductance meaning that they can be used safely in contact with electronics), compatibility with dissolved materials, low surface tension and viscosity, and high liquid density are all characteristics of HFC's, HFO's, and HFE's f-gases. Although there are alternatives in the market that may cover one or more of these product characteristics, only these f-gases give a broad enough range of each property to provide excellent value to downstream users. A comparison of fluorinated-based solvents and non-fluorinated alternatives is in Table 2 below Table 2: Comparison of fluorinated-based solvents and non-fluorinated alternatives (Chemours generated table): A.3.9.1.4. Propellants Annex A claim page 68: "Propellants are used to expel the contents of an aerosol from a canister through a nozzle, in products such as deodorants and hair sprays. Technical propellants are used for industrial uses for items such as lubricant sprays, dusters, cleaners, safety horns, degreasers, cold sprays, and paints. Propellants used in medical applications like MDI (Metered Dose Inhalers) are covered in section A.3.10. Liquified compressed gases are widely used as propellants, as they maintain a relatively constant pressure as the contents are dispensed, maintaining consistent droplet size and spray rate which may be required for technical aerosols. In contrast, compressed gases, such as carbon dioxide, cannot produce a consistent particle size and spray rate, thereby limiting their applicability, with performance falling as the contents of a can are used up and pressure within the can falls. Where a non-flammable propellant is required, HFOs are often used, alone or in a propellant blend." Chemours' Response: Fluorochemicals, primarily HFC-152a, are used to reduce the VOC content of formulations in order to reduce the emission of hydrocarbons and improve air quality. HFC-152a is modestly flammable and out of scope. HFC-134a is generally chosen when flammability is of concern in the aerosol application. A.3.9.1.6. Fire suppressants Annex A claim page 68-69 : Main sub-use categories assessed: Total flooding systems Local streaming agents Fire-fighting foams are not part of this assessment. They are covered in a separate restriction proposal. In the present assessment only clean fire suppressing agents, which are not foams, are included. ANNEX XV RESTRICTION REPORT - Per- and polyfluoroalkyl substances (PFASs) 69 Fluorinated gases (e.g. HFC-125 and HFC-227ea) are used for fire protection purposes where their main advantage is that they are `clean', non-conductive to electricity (i.e., have good dielectric properties) and are considered safe for humans to breathe at the concentrations used. In this context, `clean' refers to the ability of the fire suppressant to not leave non-volatile residues after discharge, i.e., avoid the potential damage caused by conventional extinguishing agents. This means that fluorinated gas fire suppressants occupy a niche market, when there is a need to protect items that otherwise would be damaged by a fire extinguishing agent, and in enclosed spaces where some other fire suppressants would pose a risk to human health. Fire suppressants may be divided into total flooding agents and local streaming agents. Areas of use include portable and fixed aircraft fire protection systems (e.g. engine, auxiliary power units and cargo compartments), as well as specific risk situations (e.g. clean-room protection, electronic, IT- and control room installations mainly at critical infrastructures) including the defense sector. 2-BTP (CH2=CBrCF3) is a frequently applied substance for fire suppression. The substance is a halogenated clean agent (HCA) used as halon replacement agent in handheld extinguishers onboard aircraft. Some fluoroketones, (e.g. FK-5-1-12 (CF3CF(CF3)C(=O)CF2CF3), are also introduced as a third-generation fire suppressant. Clean fluorinated gas fire suppressants may also be used in archives and museums with paper archives, historical documents, priceless works of art and antiquities where other fire protection fluids cannot be used. Fluorinated gas fire suppressants are specifically used for several military applications, e.g. in engine- and crew compartment systems on army ground vehicles (e.g. HFC-236fa) and in fixed systems protecting flight simulators and command centres. In combat the soldiers have very limited possibilities to leave the vehicle and are therefore exposed to the extinguishing media. Chemours' Response: Please consider our reviewed version of section 3.9.1.6 Main sub-use categories assessed: Total flooding systems Local streaming agents Fire-fighting foams are not part of this assessment. They are covered in a separate restriction proposal. In the present assessment only clean fire suppressing agents, which are not foams, are included. The principle fluorinated gases used for fire protection purposes are HFC-125, HFC-236fa, FK-5-1-12, and HFC-227ea and 2-BTP. Their main advantage is that they are `clean', non-conductive to electricity (i.e., have good dielectric properties) fast acting, and are considered safe for humans to breathe at the concentrations used. Fluorinated gas fire suppressants occupy a niche market, where there is a need to protect items that otherwise would be damaged by a fire extinguishing agent itself or a slow extinguishment process, and in enclosed spaces where some other fire suppressants would pose a risk to human health and safety. These niche applications have unique challenges to providing fire and life safety. Commercial aviation, for example, still uses ODP agents H1301 and H1211 while working toward migrating to alternative technologies for more than 30 years with very limited success. To date the LAVEX systems have successfully migrated from halon to HFC-236fa (in scope) and HFC-227ea (in scope) and the in-cabin portable fire extinguishers are in the process of migrating to 2-BTP (in scope). This use is indicative of the challenges of finding suitable alternatives to legacy products for fire applications as the larger engine nacelle, cargo bay, and APU uses in aircraft remain with halon. Each protected hazard using a chemical clean agent provides its own unique set of circumstances and issues which need to be overcome in order to use an alternative agent or technology. While some uses may migrate to alternative technologies with modest challenges, others cannot - and are still trying to migrate to the last generation of fluorochemical agents without significant success. Even within homologous categories such as commercial aviation, the ability to transition to alternatives varies significantly depending on the unique requirements of individual protected hazards. The diversity of applications and the unique requirements of individual facilities makes it very challenging to claim that an alternative, non-PFAS technology, is viable across a general use category - such as "Data Centers" or "Control Rooms", or "Commercial Aviation". Each Data center, for example, will have a different criteria or requirement for the fire protection technology selected. Clean Agent fire systems are not normally required by code. They are an added expense to mitigate an unacceptable risk situation to the potential loss of a site in the event of fire. This risk analysis includes high value assets and business continuity, but also societal impact and security with the loss of critical services in the event of fire. In the interconnected internet of things (IOT) world, loss of telecommunications and internet access can pose significant societal impacts. This makes the grouping of applications for regulation within the category very difficult. In some cases, the assets within the protected hazard are critical, in others it is the function of those assets within the protected space that are of value, not necessarily the assets themselves. Note that 2-BTP (CH2=CBrCF3) is NOT a frequently applied substance for fire suppression. Currently it is only used in aviation portables. Volumes are very limited, and it is ineffective as an alternative for total flood systems. One fluoroketone, FK 5-1-12, has proven successful in protecting some critical applications formerly protected with Halons or HFCs. The industry has been researching fire extinguishing molecules and technologies since the late 1980s, with very limited success commercializing alternatives due to the challenging balance of performance/extinguishment, safety for people and property, and environmental acceptability. Military vehicles (where both HFC-227ea and HFC-236fa are used), have significant performance requirements to extinguish an explosion within an occupied crew space within milliseconds to protect the crew. This is an extreme, but proven effective, lifesaving application indicative of the necessity of the current products and the challenges faced with finding and implementing alternatives. As such, an extended derogation for fluorinated fire extinguishants needs to be considered to continue fire and life safety protection for many critical applications. A.3.9.1.7. Minor uses Annex A claim page 70: "Immersion cooling is a method for cooling data centre IT hardware, including 5G network components, by directly immersing the hardware in a non-electrically conductive (Dielectric) fluorochemical liquid. The heat generated by the electronic components is directly transferred to the fluid to directly cool the components. This reduces the need for interface materials, heat sinks, fans, shrouds, sheet metal and other components that are common in traditional cooling methods. Immersion cooling systems are also simple to operate, and require minimal components for efficient operations This application is considered in detail in section A.3.12 on electric/electronic equipment." Chemours' Response: Extensive information on immersion cooling is provided in the Attachment 2, Immersion cooling application submission especially the focus on two-phase immersion cooling, which is not mentioned the applications in Annex A, section 3.9.17. In addition to the expanded description of the immersion cooling technologies, Chemours requests that ECHA considers broadening the definition of this application to include electronics, E-powertrain, and batteries as this definition would be inclusive to all applications considering this technology. In the case of automotive, multiple automotive manufactures are evaluating immersion cooling for electronics and battery cooling as a means to increase the efficiency and the overall range of the vehicle. Additionally, Tesla recently announced that they are considering immersion cooling technologies to enable faster charging by submerging the electrical components in the charging stations to improve the efficiency of the stations themselves.18 A.3.9.2. Volumes Annex A claim page 70-71: Table A.36 p70 is showing "Yearly total volume of HFCs and PFCs in EEA per main use category." 18 https://www.inverse.com/gear/tesla-liquid-cooled-charging-cybertruck-semi-fast-charge Chemours' Response: Table A.36 is reporting data from EEA GHG inventory 2018 submitted 2020 v2, and specifically Table2(II)BHs2.19 The reference year of the data is not indicated in Table A.36, and numbers need to be expressed in the correct measuring units (tons for stock for example). The reported numbers include non-PFAS fluorinated gases which should be deducted. Table A.36 should be reviewed accordingly. It should also be indicated when data is used that contains substances that are not considered as PFAS, which represent more than 25% of the total supply according to figure A.19 on page 76 of annex A. Removing HFC-32 and HFC-152a from the figures in table Table A.36, the volumes would look as follow: 19 CRF tables EU-27+UK. European Environment Agency. (2020, May 28). https://www.eea.europa.eu/publications/european-union-greenhouse-gas-inventory-2020/eu-27-crftables.zip/view Table A.36. without non-PFAS (HFC32 and HFC-152a) Total Manufactured products 6886,20 t/y Commerical refrigeration Stocks 85259,77 t Decommissioning 5472,47 t/y Manufactured products 119,53 t/y Domestic Refrigeration Stocks 4395,01 t Decommissioning 668,57 t/y Manufactured products 1805,80 t/y Industrial Refrigeration Stocks 31415,51 t Decommissioning 1155,25 t/y Manufactured products 891,38 t/y Transport refrigeration Stocks 8659,84 t Decommissioning 198,99 t/y Mobile Air Conditioning Manufactured products Stocks 5176,39 t/y 113350,71 t Decommissioning 4414,69 t/y Manufactured products 4214,15 t/y Stationary Air Conditionig Stocks 92228,67 t Decommissioning 4840,94 t/y Manufactured products 3525,07 t/y Foam blowing agent (closed cell) Stocks 45939,17 t Decommissioning 169,64 t/y Manufactured products 6,30 t/y Foam blowing agent (open cell) Stocks 7450,04 t Decommissioning 0,00 t/y Fire protection Manufactured products 862,86 t/y Stocks 20145,37 t Decommissioning 208,10 t/y Manufactured products 144,92 t/y Aerosols (non-MDI) Stocks 554,73 t Decommissioning 0,00 t/y Manufactured products 0,00 t/y Solvents Stocks 0,00 t Decommissioning 0,00 t/y Manufactured products 0 t/y Other Stocks 0 t Decommissioning 0 t/y Total Manufactured products Stocks 23632,61 t/y 409398,82 t Decommissioning 17128,66 t/y Annex A claim page 78: "...one estimate was provided by the Environmental Investigation Agency which estimated that 16.3 million tons CO2 equivalents of bulk HFCs were illegally placed on the EU market in 2018 (EIA, 2019)." Chemours' Response: Another data point from the EFCTC (European Fluorocarbon Technical Committee20, a sector group of CEFIC estimated that the illegal trade of HFCs could represent up to 31 million tons CO2 equivalent (MtCO2e) in 2019.21 This could represent up to 30% of the legal HFC market. A3.10 Medical devices A.3.10.1.11. Propellants in Metered Dose Inhalers (MDI) Annex A claim page 83: "Fluorinated gases are also applied in metered dose inhalers (MDI) where they act as a propellant for the active pharmaceutical ingredient (API). In 1987, the Montreal Protocol was signed and called for the elimination of CFC propellants. Metered-Dose Inhalers (MDIs) are typically used for the treatment of asthma and other respiratory conditions. These devices are regulated under the Aerosol Dispenser Directive. MDI, nasal sprays and nebulizers are used to administer pharmaceuticals directly into the lungs. This enables the achievement of high active pharmaceutical ingredient concentrations, while minimizing systemic exposure. The bestknown application of MDI is the treatment of patients with COPD or asthma. Additionally, treatment of cystic fibrosis, chronic lung infections, influenza, osteoporosis, pulmonary hypertension has been reported (Stein and Thiel, 2017).22 The number of pMDI (pressured MDI) is 20 million per year." Chemours' Response: In addition to the "MDI, nasal sprays and nebulizers" listed as available technologies the largest practical alternative to MDI aerosol inhalers are DPI, Dry Powder Inhalers. DPI technologies are the primary alternative option for MDIs but have multiple limitations and drawbacks in use, most critically the requirement of a patient to forcefully breath in the medicament - a challenge to many patients with respiratory disease. Because of the wide range of DPI designs, there are challenges in developing information and instructions in support of the device, and the functionality of the device might not last as long as an MDI. DPIs are also more susceptible to contamination because of their design and drug delivery method whereas MDIs are not; DPIs also contain lactose, though in small amounts. DPIs are generally more expensive to produce as well.23 More detail is available in the Attachment 2, Medical Devices application submission. A.3.10.1.16 Fluorinated gases Additional clarification to A.3.10.1.16 Fluorinated gases: 20 https://www.fluorocarbons.org 21 https://www.stopillegalcooling.eu/wp-content/uploads/2022/06/Press-release_Oxera_EN_FINAL1.pdf?_gl=1*rlka00*_ga*NTA1MjM5OTM3LjE2NzkzMTg0Njg.*_up*MQ 22 Stein SW, Thiel CG. The History of Therapeutic Aerosols: A Chronological Review. J Aerosol Med Pulm Drug Deliv. 2017 Feb;30(1):20-41. https://doi.org/10.1089/jamp.2016.1297. Epub 2016 Oct 17. PMID: 27748638; PMCID: PMC5278812. 23 Comparative studies on metered dose inhalers and dry powder inhalers: Use inhalers. Comparative studies on Metered dose inhalers and Dry powder inhalers | Use Inhalers. (n.d.). https://use-inhalers.com/comparativestudies-on-mdi-and-dpi Chemours' Response : From the UNEP MCTOC 2022 Report24: Concerning Orally inhaled CFC propellants for metered dose inhalers: Under the Montreal Protocol, the use of CFCs as propellants for pMDIs was successfully phased out worldwide without significant adverse impact to medical care. Pharmaceutical CFCs were replaced with HFC propellants in pMDIs; HFC-134a and to a lesser extent HFC-227ea. This transition required more than a decade to execute - with a well-coordinated, global effort, managed under IPAC. With the Kigali Amendment, the production and consumption of HFCs listed in Annex F are scheduled to be phased down. Annex F HFCs include HFC-134a, HFC-227ea and HFC-152a. The current MDI transition has been more independent and less well coordinated across the user community. This makes a quick transition challenging, leaving patient treatment at risk. While the industry is moving quickly toward a transition to HFC-152a, depending on the EiF date, it will be challenging for all formulations and variations to fully complete the development and approval process within the time allotted. Annex A: A.3.10.1.18: PFAS liquids for cleaning and heat transfer fluids: engineered fluids Annex A claim page 87: "Cleaning applications include the cleaning of metal and plastic parts, such as orthopaedic, dental and spinal implants, artificial hearts, heart valves, catheters, needles and stents. Often engineered (fluorinated) fluids are used for cleaning and rinsing. The mentioned solvents are intended for industrial use only and are not intended for use as a medical device or drug. Perfluorinated engineered fluids are replacements for n-propyl bromide, trichloroethylene (TCE), ozonedepleting solvents such as HCFC-225 and HCFC-141b, and HFCs with high global warming potential. For heat transfer in medical equipment (e.g. surgical lasers) and laboratory diagnostic devices other PFAS fluids like 1-methoxyheptafluoropropane and 3-ethoxyperfluoro(2-methylhexane) are used. An overview of to what extent specific PFASs (types) are used throughout the medical devices industry is presented in Figure A.20. This information was provided by the members of Spectaris (a German industry association for the high-tech business sector). An overview of other uses is mentioned in Table A.106." Chemours' Response: In section 3.10.1.18, cleaning and heat transfer applications are mentioned for medical devices. Chemours requests to add carrier fluid applications to this list. During the manufacturing process, fluorinated solvents are used to as a carrier to dilute fluorinated lubricants and silicone. Because the fluorinated solvent is volatile, there is no impact to post treatment of the medical devices in product sterilization. Historically, flammable aliphatic and aromatic organic solvents were commonly used for dilution and deposition. Fluorinated solvents are also a drop-in replacement for chlorinated solvents, which are increasingly subjected to environmental regulations due to its high ozone depleting potential and other health hazards. Fluorinated solvents offer high solvency, evaporation rate, low surface tension, and nonflammable properties enable desirable and even coating of deposition product with negligible residue on the medical device.25 24 https://ozone.unep.org/system/files/documents/MCTOC-Assessment-Report-2022.pdf 25 XF Medical Lubricants. (n.d.-e). https://www.microcare.com/MicroCare/media/TDS/XF-Medical-Lubricants-USTDS.pdf?ext=.pdf The main application for Chemours Solvents in medical devices is for implantable devices. Our solvents are used as carrier fluids that evaporate during the manufacturing process with minimal residue and no solvent left behind. The cleaning of these parts (Pacemaker controls, valves, implants, etc.) is critical for the health of the patient and cannot be accomplished with alternatives such as water alone. As referenced in A.3.10.1.5, our products can also be considered for heat transfer or engineered fluids. A3.10.1.18 is only focusing on cleaning. Chemours Solvents would likely fall under F-gases or PFAS type unknown. As mentioned, Perfluorinated engineered fluids are replacements for n-propyl bromide, trichloroethylene (TCE), ozone-depleting solvents such as HCFC-225 and HCFC-141b, and HFCs with high global warming potential. Chemours HFO Solvents are replacements for HFCs and Perfluorinated engineering fluids by offering a more attractive environmental profile and better performance. The other alternatives listed (nPB26 and TCE27) both have documents health hazards and toxicity concerns, which make them not viable alternatives. A 3.10.2. Volumes A.3.10.2.2 Fluorinated gases Annex A claim page 88 : "A total approximately 33,000 (midpoint) fluorinated gases are used in industrial processes related to medical devices like MDI's, medical lasers according to the ECHA database. Disaggregation of the tonnage to medical devices in scope is not always possible, however. On top of the mentioned tonnage, fluorinated gases are used in exempted uses such as anaesthetics, contrast media and pharmaceutical use which is exempted as well (HCWH, 2019). Three gases are responsible for 99.9% of the medical fluorinated gases reported (based on data from ECHA search and response to the CfE). In table Table A.104 in the appendix the greenhouse warming potential of these gases is listed as well as the importance of HFC-134a. HFC-134a is the most used gas, followed by HFC227ea and HFE-152a (HFE-152a is outside scope). Generic worldwide use of these three main medical gases is mentioned in (Booten et al., 2020). The volume of fluorinated gases for metered dose inhalers (MDI's) has been estimated in three different ways: production based on stakeholder information, ECHA database information, and MDI sales data/a report of Health Care Without Harm (HCWH, 2019). Volumes ranged between: 6 000 t/y (stakeholders); 400 t/y (HCWH and MDI sales data); 15 000 - >30 000 t/y (ECHA volumes, including volumes for export: amongst others HFC-134a: 12 000 - 20 000 t/y, HFC-227ea >3 000 t/y and HFC-152a 650- 6 500 t/y. All numbers including production for export). Stakeholder information (6 000 t/y) was used for impact assessment." 26 5. unreasonable risk determination - U.S. environmental protection agency. (n.d.-a). https://www.epa.gov/system/files/documents/2022-12/1-BP_Final%20Revised%20RD_12-12-22.pdf 27 5. unreasonable risk determination - U.S. environmental protection agency. (n.d.-a). https://www.epa.gov/system/files/documents/2023-01/TCE_Final%20Revised%20RD_12-21-22-FINAL-v2.pdf Chemours' Response: The UNEP MCTOC MDI global volume estimates for MDI propellants is approximately 11 500 mt/yr with a significant portion split between manufacturing in the US and Europe. As such, 6000mt/y for European production is within scale. Note production of MDI inhalers is distributed globally and a constraint imposed on European production will have significant repercussions in countries and patients around the world.28 A 3.11 Transport A.3.11.1. Uses Annex A claim page 94: HVACR-systems in transportation : "Use of fluorinated gases in the various HVACR-systems in transport vehicles for passenger cabin air conditioning or transport refrigeration" Chemours' Response: Chemours finds that the examples listed are not representative and that there is repetition. Chemours suggests replacing by: "Use of fluorinated gases in the various HVACR-systems in refrigerated transport vehicles, railways and marine containers, for cabin comfort air conditioning and for EV thermal management (electronics, power trains, batteries)." Furthermore, Chemours suggests removing cleaning fluids and blowing agents, which should not be included here based on their uses. Annex A claim page 94: "Special heat transfer fluids (e.g. Methoxyheptafluoropropanes) for the immersion-cooling/heating of electronic equipment." Chemours' Response: Methoxyheptafluoropropanes is a potential fluid option for immersion-cooling/heating of electronic equipment, and there are other options such as HFOs (for example HFO-1336mzzZ) featuring Low viscosity at very low temperatures and excellent dielectric properties. A 3.12 Electronics and semiconductors A.3.12.1. Uses Annex A claim page 104: Table A.45: PFAS properties relevant to the electronics and semiconductor industry - Literature and publicly available sources, complemented by a stakeholder. "Heat resistance, low dielectric constant, clearness, plasma resistance, high photosensitivity, ability to generate acids, low surface tension, Marangoni effect, low refractive index, acidic, non-reactive, stable, 28 Montreal protocol on substances that deplete the ozone layer December 2022. (n.d.-c). https://ozone.unep.org/system/files/documents/MCTOC-Assessment-Report-2022.pdf non-corrosive, temperature uniformity, generation for reactive oxygen/fluoride species, chemical resistance, high purity, anti-adhesion, insulation, barrier properties, thermal stability." Chemours' Response: In Table A.45, additional properties for semiconductor would include wide operating temperature ranges, low freezing point and low viscosity. Annex A claim page 105: Table A.47 Table A.47. Uses of PFASs in the production process of electronics and semiconductor products and components. Complemented by a stakeholder. "Heat transfer fluid | Submersion cooling, chemical vapour deposition" "Etching" Chemours' Response: In Table A.47, in the column "used as/for" Heat Transfer Fluid there are critical applications missing. The applications to also include are: Chemical Vapor Deposition, Physical Vapor Deposition and Photolithography. In the column "used as/for" etching, these processes should also be included for the semiconductor industry. Annex A claim page 106: Table A.48: Uses and properties of PFASs in the electronics industry identified by stakeholders. Chemours' Response: In section 3.12.2, table A.48 page 106. In the Coating of Electronic Components use category, two additional carrier fluids should be included, specifically 1,1,1,2,2,3,4,5,5,5- Decafluoropentane (HFC4310mee) and Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO-1336mzzZ). Table A.48 page 109. information on properties and uses as heat transfer fluids are missing. Chemours recommends that the following information be added for single-phase and two-phase applications to complete the proposed table for further review by industry. In the properties column the following points should be included: Normal Boiling Point (NBP): the normal boiling point (NBP), which will distinguish between a fluid used in single-phase (NBPs>~150degC) or two-phase immersion cooling (NBPs between 4060degC) Flammability: Flammability is used to define the risk associated with a material when it comes to ignition and flame propagation properties. Dielectric Constant and dissipation factor (or loss tangent): Dielectric constant and dissipation factor are measurements used to qualify a fluids electrical insulating properties. These properties are critical as they directly impact signal integrity (or signal strength) when electronic components such as printed circuit boards and cables are immersed in the fluid. Breakdown voltage and volume resistivity: additional dielectric properties that impact the creepage distance and electrical resistance imposed by the fluid, respectively. Thermal Stability: Thermal stability is a measure used to define the operational range of a fluid before the fluid degrades and breaks down. This measure is critical for design as it advises when acids and other corrosive materials can form in a fluid.  Global Warming Potential (GWP): Global warming potential as defined by the AR4 standard is used to determine any direct emission potential from a fluid. Ozone Depletion Potential (ODP): Ozone depletion potential is the measure used to determine if any detrimental ozone impacts are present with the fluid. Material compatibility: Material compatibility is used to determine if there are any materials that would degrade when using a working fluid. This impacts not only fluid selection but component design. In addition to the above properties, the following applications should be added to the cell "Area of use/ application(s)" under heat transfer fluids: Cooling of IT Infrastructure equipment Cooling of telecommunications infrastructure equipment Electronic cooling Battery cooling, or cooling of energy storage devices Cable and electrical transmission cooling Electric vehicle powertrain (e-powertrain) Furthermore the Heat Transfer fluids Use category only mentions immersion cooling. Chemours recommends adding an additional sub use under the "Heat Transfer" use category--thermal management for semiconductor process. The area of use/application includes: Chemical Vapor Deposition, Physical Chemical Deposition, Photolithography, etching, and test processes. Additionally, Methoxytridecafluoroheptene isomers (MPHE) and Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO-1336mzzZ) should be listed as an example product in this section. Their properties include nonflammable, nonconductive, high dielectric strength, low freezing point, high thermal stability, low viscosity, and low surface tension. Table A.48 page 110. The Solvent use category should include 1,1,1,2,2,3,4,5,5,5- Decafluoropentane (HFC-4310mee), Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO-1336mzzZ) and Methoxytridecafluoroheptene isomers (MPHE). Their properties include nonflammability, targeted solubility for contaminants for lubricants of interest, quick drying, low residue, low surface tension, low toxicity, low viscosity, high density, and nonconductive. Table A.48 page 110. The aerosol/solvent cleaning of electronic components use category should include 1,1,1,2,2,3,4,5,5,5- Decafluoropentane (HFC-4310mee), Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO1336mzzZ) and Methoxytridecafluoro-heptene isomers (MPHE). Their properties include nonflammability, and inerts other flammable solvents. Annex A claim page 112: Table A.49: Uses and properties of PFASs in the semiconductor industry identified by stakeholders. Chemours' Response: This table covers semiconductor applications specifically. As a heat transfer fluid, Methoxytridecafluoroheptene isomers (MPHE) and Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO-1336mzzZ) should be included as an example. Additionally, for Thermal testing of semiconductor devices, Methoxytridecafluoro-heptene isomers (MPHE) and Cis-1,1,1,4,4,4-Hexafluoro-2- butene (HFO-1336mzzZ) should also be included as examples. A 3.18 Waste A.3.18.1. Introduction Annex A claim page 154: "Generally, waste streams for small residential appliances such as small air conditioning units, differs from the end-of life treatment for large commercial and industrial systems: For the smaller appliances collection, storage and treatment is organized under WEEE regulation. For the larger systems certified technical personnel is needed for the mandatory recovery of the Fluorinated gases according to the Fgas Regulation." Chemours' Response: Even for small residential appliances in EU, there is a requirement that F-gas certification is needed in order to install, maintain and dispose of F-gas containing equipment to be allowed to touch the installation. Recovery and Destruction of F-gases is specifically regulated by the EU No 517/2014. Article 8, "Recovery", indicates:, "Operators of stationary equipment or refrigeration units of trucks and trailers shall ensure that the recovery is carried out by natural persons that hold the relevant certificates provided for by Article 10, so that those gases are recycled, reclaimed or destroyed."29 This applies to cooling circuits of: stationary refrigeration, air-conditioning and heat pump equipment; refrigeration units of trucks and trailers; stationary equipment that contains F-gases based solvents; stationary fire protection equipment; and stationary electrical switchgear.. Article 19, "Reporting", indicates that "quantities of each substance listed that have been recycled, reclaimed and destroyed shall be reported."30 A detailed description of the end-of-life provisions established by the F-gas Regulation (EU 517/2014) is provided in Attachment 1, Chapter 1, section 1.5 of this submission. A.3.18.2.5. End-of-Life-Vehicles (ELV) Annex A claim page 160: Chemours' Response: General comment on ELV for Mobile air conditioning applications: The main refrigerants in use today for MAC applications (R-134a and R-1234yf) can be reused many times at the end of the life of the vehicle. The refrigerant should be recovered so it can be reused.31 Chemours does not currently have specific data to quantify the amount of vehicles for recycling in the EU and for export, nor how big this opportunity may be. The 12 million vehicles leaving the road would equate to roughly 6 000MT of refrigerant for potential reuse (0.5kg/vehicle). Appendix A.2. Manufacture, Import, and Uses Annex A claim pages 166 - 175: Table A.70: PFAS from the OECD and REACH registry database combined. Chemours' Response: The use of REACH registration midpoint volumes to estimate production and import is less robust than the official F-Gas reporting and is misleading: it is accounting for both manufacturing and import. Registered tonnage band do not represent actual production or imports and is not reflecting real market size.. Confronting mandatory EEA report official data (15 000 tons/yr32) with 176 548 tons/yr mentioned in Table A.70 would mean c.a. 160 000 tons/yr is not reported by manufacturing companies in EU.33 As a further remark, only official F-gas reports should be accounted as trusted source of information. Annex A claim page 176: Table A.73: A summary of annual imports of PFAS chemicals from third countries into EU-27. Chemours' Response: Table A.73 is reporting inconsistent data for fluorinated gas volumes. The most reliable source of data on Total EU import is the EEA Report 2020 Table A5.5.34 32 Fluorinated greenhouse gases 2021. European Environment Agency. (2021, Dec 8) https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2021/fluorinated-greenhouse-gases2021-annex/view table 1 33 Registered substances. ECHA. (n.d.-a). https://echa.europa.eu/information-on-chemicals/registered-substances 34 Fluorinated greenhouse gases 2020. European Environment Agency. (2020, December 1). https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2020 Annex A claim page 177: Table A.74: A summary of annual exports of PFAS chemicals from EU-27 into third countries (t). Chemours' Response: Table A.74 is reporting inconsistent data for fluorinated gas volumes. The most reliable source of data on EU bulk exports is the EEA Report 2020 Table A5.13. There is no data available to account for export in products and equipment.35 Appendix A.3.9 Applications of fluorinated gases Annex A claim pages 248 - 259: Table A.95 : Fluorinated gases currently in commercial use for Heating, Ventilation, Air Conditioning and Refrigeration (HVACR) and other uses - Organised by HFC Code (Source: Stakeholder consultation and literature review carried out during the development of the restriction proposal). Chemours' Response: Chemours recommends the following for each substance below: HFC- 43-10mee: "Heat Transfer Fluid" should be added as a sub-use in Solvents. 35 Fluorinated greenhouse gases 2020. European Environment Agency. (2020, December 1). https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2020  HFC-134a: Chemours is unaware of the use of HFC-134a as a propellant in Consumer Products. There is use in some aerosol applications where flammability of the propellant is a safety issue. These are mostly in technical aerosol lubricants and cleaners. HFC-152: consider adding catalyst regeneration applications. HFC-245fa: Should consider adding Working Fluid for industrial applications, Organic Rankine Cycles for energy recovery, Secondary fluid for Commercial Refrigeration. HCFC-124 has very limited uses in Fire. HFO-1234zeE: under "specific use": Not only data centers: working fluid in chillers and heat pumps for several industrial applications. HFO-1336mzzE: under refrigeration and heat pumps general use should add Waste Heat Recovery Applications with High Temperature Heat Pumps and Organic Rakine Cycle; HFO-1336mzzZ: In the General Use "Solvent" is missing "Heat transfer" as an application for HFO- 1336mzzZ: Chemours is unaware of the use of HFO-1336mzzZ in commercial propellants.HCFO1223yd: under specific use, we should add centrifugal chillers. HCFO-1224ydZ: under specific use, should add Organic Rankine Cycle (ORC) applications. MPHE, SionTM: We recommend to remove SionTM as this is not an accurate name for this product, there is also a need to add "heat transfer fluid" as a sub-use. FK 5-1-12 is used for total flood primarily, and smaller limited uses are in local streaming applications. Chemours recommends including fluoroketones, Hydrofluroofelins, and perflorocarbon as both substances are used in immersion cooling applications today and would complement the current list provided in this section. R-452A: under Sub-use, should add Industrial direct-expansion refrigeration systems. Under Specific use should add Refrigerated Trucks and Trailers. R-452B: under Specific use: can add Air/Water Chillers for commercial applications. R-515B: under Specific use: add High ambient temperature applications. Annex A claim page 263: Table A.97 : Intended applications of EU-28 total supply of fluorinated gases. Source data reported in the F-gas Report (EEA, 2020). Chemours' Response: These historical values seem accurate and commensurate with our experience over a similar period. They are an argument against the 5.9% growth rate for chemical fire suppression products in Europe mentioned in Annex E section E.2.8.1. The 5.9% value is an estimate based on overall industry growth rate. The implementation of the F-Gas regulations has seen substantial reduction in the use of fluorinated extinguishing agents in Europe as supported by the historical data in this table, with an average 35% loss year on year from 2015 through 2019. From this, a reasonable expectation is that the supply of fluorochemical extinguishing agent to Europe will be flat to further declines year on year as the F-Gas regulations continue to constrain regulated products. As discussed elsewhere, the use of fluorinated extinguishing agents is a critical requirement for some applications. The existing F-Gas regulations are working to transition all non-critical uses to other technologies; as such an extended derogation for those remaining fire applications, where no other option of transition is feasible, is required. Future volumes for the derogated fire segment will continue to be flat year on year to decreasing, driven by only those applications and facilities where the requirement for a fluorochemical clean agent is critically necessary. Annex A claim page 264: Table A.98: Estimation of quantities of Hydrofluoroolefins Used , as a Proportion of F-gases in Products and Equipment for EU-28. Source: EEA 2020 Chemours' Response: While this picture could be correct for the entire industry, it should however be reminded that the portion of HFO differs heavily from one market to another. The mobile AC market is using predominantly R-1234yf and the portion there is >90%. The HFO percentage is correct. It is the result of combining data in EEA Report 2020 in Table A.5.21 and Table A.5.17.36 36 Fluorinated greenhouse gases 2020. European Environment Agency. (2020, Dec 1) https://www.eea.europa.eu/publications/fluorinated-greenhouse-gases-2020 Appendix A.3.10 Medical devices Annex A claim page 269: Table A.104: Estimated fluorinated gas greenhouse warming potential (GWP) based on reported volumes of medical gases in the EU (responses to CfE and ECHA database) and assuming an emission factor of 0.1%. Chemours' Response: In table A.104, R134a in medical gases looks very high though this is for all possible medical uses. UNEP MCTOC estimates for MDI use globally are approximately 10 600 tons (in 2018) for HFC-134a; HFC-227ea approximately 900 tons. There may be some additional 134a Medical uses beyond MDI, but the volumes are unlikely to be significant. 25 000 tons just for the EU is 2.5X UNEPs global estimate. There are no known medical uses for HFC-227ea other than pMDIs, and the global production is approximately 9001000 tons/yr for that application. A volume of 3 000 tons is significantly excessive by 3x of the global volumes.37 In table A.104, HFC-4310mee volume is listed at 5.5 MT/year. This is too high and the volume likely takes account total volume of blended products. Amount of HFC-4310mee in its pure form is estimated to be lower. Additionally MPHE is not listed on this table. 37 Ohnishi, K. et al. (2018) Medical and Chemical Technical Options Committee: 2018 Assessment Report, UN Environment Programme Ozone Secretariat, pp. 12. https://ozone.unep.org/sites/default/files/2019-04/MCTOCAssessment-Report-2018.pdf Annex III: Comments on the analytical methods described by the Dossier Submitters in Annex E of the PFAS restriction dossier Comments to section E.4.1.4 of the restriction proposal dossier, pages 521 - 532 on analytical methods Page 521 reports a general introduction to analytical methods for PFAS detection. Application to our substances in general of the analytical methods listed is uncertain, as it will depend on the specific substance. With sufficient effort and availability of reference samples the analytical methods in this paragraph would likely be applicable. Page 525 refers to mass spectrometry as a suitable technique for PFAS detection. The method CEN/TS 1596838 is not applicable to F-gases, it is specifically designed for determination of perfluorooctanesulfonate (PFOS) in concentrated extracts from coated and impregnated solid articles, liquids and fire extinguishing foams. All substances are either solid or high boiling liquids as shown in Table 3 below (page 8 in this method). Table 3 - Analytes determinable by method CEN/TS 15968 Page 525 refers to different analytical standards available for the monitoring of PFAS in environmental samples. As a general comment, these methods/standards are not applicable to F-gases. They are designed and validated for the analysis of non-volatile PFAS in various aqueous media such as drinking 38 PD CEN/TS (no date) PD Cen/TS 15968:2010, ANSI Webstore. Available at: https://webstore.ansi.org/standards/bsi/pdcents159682010?gclid=EAIaIQobChMIgf7S_7a4_QIVi__jBx2i0QAbEAAY ASAAEgKt1PD_BwE. water, wastewater, and body fluids. All of the examples provided in these methods/standards are acids (carboxylic, sulfonic) and sulfonamide derivatives. A more detailed assessment is provided below: - EPA OTM-45: Specifically states boiling point > 100 C. All of the examples presented in the method are acids (carboxylic, sulfonic) and sulfonamide derivatives. Applicability to F-gases is unclear and would need to be determined. - EPA Method 537: The method is designed for PFAS "Alkyl Acids" in drinking water samples. Only carboxylic and sulfonic acids are given as examples. Sample procedures described are likely not directly applicable to F-gases but possibly could be modified to analyze them. The first sentence of the Scope and Application states "This is a liquid chromatography/tandem mass spectrometry (LC/MS/MS) method for the determination of selected perfluorinated alkyl acids (PFAAs) in drinking water. - EPA Method 533: The method is designed for PFAS in drinking water samples. Only carboxylic and sulfonic acids are given as examples. Sample procedures described are likely not directly applicable to F-gases but possibly could be modified to analyze them. The first sentence of the Scope and Application states "This is a solid phase extraction (SPE) liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for the determination of select per- and polyfluoroalkyl substances (PFAS) in drinking water." - EPA Draft Method 1633: The method is designed for PFAS in aqueous, solid, biosolid, and tissue samples. Examples are limited to carboxylic and sulfonic acids. Sample procedures described are likely not directly applicable to F-gases but possibly could be modified to analyze them. - EPA Method 8327 Limited to samples obtained from water matrices. All 24 PFAS analytes that have been evaluated with this method are carboxylic or sulfonic acids. The Scope and Application states "This determinative method may also be applicable to other PFAS target compounds and other matrices, provided that the laboratory can demonstrate adequate performance (refer to Sec. 9.0 or project-specific acceptance criteria) using representative sample matrices." - EPA Response to Public Comments on SW-846 Update VII, Phase 2 - Method 8327 for Per- and Polyfluoroalkyl Substances (PFAS) by Liquid Chromatography/Tandem Mass Spectrometry (LC/MS/MS): This document provides EPA responses to public comments this method, no new methods or applications of the method. - ISO 21675: This method is designed to analyze specific PFAS substances as listed in Table 1. All of these analytes are straight-chain fluorinated acids (carboxylic, sulfonic) or sulfonamides, and the substrate is water (drinking, natural, wastewater). The Scope (page 7) states "The applicability of the method to further substances, not listed in Table 1, or to further types of waters not excluded, but is intended to be validated separately for each individual case." Therefore, the method would need to be adapted and validated before it could be applied to F-gases. - ISO 25101: This method is very specific and limited to analysis of PFOA and PFOS. Some leeway is provided for isomers of these two analytes. Extension to other PFAS or F-gases is not within the scope of this method. - ASTM D7968-17a: The Scope states that the method "covers the determination of selected polyfluorinated compounds (PFCs) in a soil matrix". The analyte examples given are limited to fluorinated acids (carboxylic and sulfonic). This method is not applicable to F-gases. - DIN 38407-42: This standard is not applicable to F-gas determination. As stated in the Scope (Section 1), the method is "for the determination of selected perfluoroalkylated substances in drinking, ground and surface water as well as treated wastewater". Table 1 lists the ten selected (e.g., specific) substances to which this method applies; the list includes 7 carboxylic acids and 3 sulfonic acids. The Scope further states that "The applicability of the method to further substances... needs to be checked on a case-by-case basis". In general. HPLC is also not designed for gas analysis. Therefore, even the method may be reworked and applied to other substances, it will not work for F-gas. Page 530, Table E.171 refers to the available analytical methods for different matrices. DIN EN 14582 method will determine F-gases, with caveats. The procedure involves total oxidative combustion of the sample, followed by analysis for halide (F-, Cl-, Br-, I-) and sulfate (SO42-). Many of the detection methods listed provide total halide content. If only fluoride is to be determined, then a technique specific for that anion must be used. The method is not specific to the individual species contributing to the fluoride measured, they are summed together. Therefore, if a sample contains one or more F-gases but also contains PFAS or inorganic fluoride the measurement provides the total fluoride off all of these components. Annex IV: Socio-Economic Analysis and Impact Assessment of a potential REACH Restriction on F-gases as PFAS, report for Chemours Thermal Specialized Solutions (TSS) [attached in confidential version] Annex V: Evaluation of Costs based on Restriction Derogations for Mobile Air Conditioning Maintenance Evaluation of Costs based on Restriction Derogations for Mobile Air Conditioning Maintenance Final Report prepared for Chemours TSS Division 13 September 2023 Evaluation of Costs based on Restriction Derogations for Mobile Air Conditioning Maintenance September 2023 Final Report Quality Assurance Project reference / title Report status Author(s) Approved for issue by Date of issue J1170/HVACR cost estimates Final Report Robert White Julianne Oakley Vithul Payangott David Carlander 1 September 2023 Document Change Record Report Version Draft Final Report 1.0 Draft Final Report 1.1 Draft Final Report 1.2 Final Report 1.3 Date 23 June 2023 07 July 2023 10 August 2023 13 September 2023 Change details Initial document. Initial comments addressed and model refinements implemented. Second round comments addressed, and final model refinements implemented. Final comments addressed. Disclaimer The views and propositions expressed herein are, unless otherwise stated, those of Risk & Policy Analysts and do not necessarily represent any official view of the TSS division of Chemours or any other organisation mentioned in this report. Recommended citation: RPA (2023): Evaluation of Costs based on Restriction Derogations for Mobile Air Conditioning Maintenance, report for Chemours Thermal Specialised Solutions (TSS), September 2023, Norwich, Norfolk, UK Table of Contents Table of Contents......................................................................................................................... 1 Glossary....................................................................................................................................... 2 1 Introduction ......................................................................................................................... 3 1.1 Proposed per and polyfluoroalkyl substances restriction dossier................................................. 3 1.2 Derogations..................................................................................................................................... 3 1.3 Structure of this report ................................................................................................................... 4 2 Scope and methodology ....................................................................................................... 5 2.1 Maintenance derogations assessed in this study ........................................................................... 5 2.2 Aim of study .................................................................................................................................... 6 3 Fgas maintenance derogation cost estimates ..................................................................... 11 3.1 Vehicle sale estimations (n1 and n2 estimates) ............................................................................. 11 3.2 Scenario 1 cost estimate ............................................................................................................... 15 3.3 Scenario 2 cost estimate ............................................................................................................... 19 3.4 Conclusions and recommendations..............................................................................................20 Annex 1 ICE vehicle sales estimates data ............................................................................... 23 Annex 2 Detailed scenario cost breakdowns .......................................................................... 25 Annex 3 Cost variable data tables.......................................................................................... 27 MAC Maintenance Derogation Impact Assessment RPA | 1 Acronym BEV BPR CO2 EiF EV Fgas HFC HFO HVACR ICE MAC PFAS RPA Glossary Description Battery electric vehicle Biocidal products regulation Carbon dioxide Entry into Force Electric vehicle Fluorinated gas Hydrofluorocarbon Hydrofluoroolefin Heating, ventilation, airconditioning, and refrigeration Internal combustion engine Mobile air conditioning Per and polyfluoroalkyl substances Risk & Policy Analysts ltd MAC Maintenance Derogation Impact Assessment RPA | 2 1 Introduction 1.1 Proposed per and polyfluoroalkyl substances restriction dossier Per and polyfluoroalkyl substances (PFAS) are a group of widely used manmade organic chemical substances containing very strong, stable carbonfluorine bonds. Five European countries, Norway, Germany, Denmark, Sweden, and The Netherlands, prepared a Proposal for a Restriction, submitted in January 20231 under the EU REACH Regulation (EC) 1907/2006 (REACH). In the Registry of Restriction Intentions of the PFAS restriction, PFAS are defined as "substances that contain at least one fully fluorinated methyl (CF3) or methylene (CF2) carbon atom, without any H/Cl/Br/I attached to it." The proposed restriction definition is similar to the Organisation for Economic Cooperation and Development (OECD) definition2, derived in 2021, which reads "PFAS are defined as fluorinated substances that contain at least one fully fluorinated methyl or methylene carbon atom (without any H/Cl/Br/I atom attached to it), i.e. with a few noted exceptions, any chemical with at least a perfluorinated methyl group (-CF3) or a perfluorinated methylene group (-CF2-) is a PFAS." This restriction will have significant effects on the TSS division of Chemours, downstream manufacturers and consumers3. This is due to Chemours TSS being a major market producer and importer of fluorinated gas (Fgas) refrigerants4, which are considered a PFAS and fall within scope of the proposed restriction. 1.2 Derogations Within the current proposed restriction are derogations for specific sectors and applications. These derogations have either already been given or are proposed to be reviewed after more evidence has been gathered via the public consultation process. Derogations that require more evidence will not be granted unless suitable evidence is gathered in the consultation. There are few exceptions that have been proposed for only PFAS used in active substances. These exceptions are related to active substances in biocidal products, in plant protection products and in human and veterinary medicinal products. The derogations that have been given or being considered for review are listed on page 4 and further explained in Table 9 of the restriction proposal. Additionally, there are derogations for nonpolymeric PFAS and derogations for fluoropolymers and perfluoropolyethers. Nonpolymeric PFAS derogations consist of 20 accepted derogations and a further 11 derogations that will be further reviewed following the ECHA public consultation. Fluoropolymers and perfluoropolyethers derogations consist of six accepted derogations and a further nine derogations that will be reviewed. For timelimited derogations there are reporting requirements 1 Registry of Restriction Intentions until outcome. PFAS Restriction proposal found here https://echa.europa.eu/registryofrestrictionintentions//dislist/details/0b0236e18663449b accessed June 2023 2 OECD. Series on Risk Management No. 61, 2021. available at https://www.oecd.org/env/ehs/risk management/seriesonriskmanagementpublicationsbynumber.htm accessed June 2023 3 See socioeconomic impacts report produced by RPA for Chemours TSS in January 2023 4 This includes HFC and HFO refrigerants for a wide variety of applications including but not limited too mobile air conditioning, commercial refrigeration, transport refrigeration and pharmaceutical applications. MAC Maintenance Derogation Impact Assessment RPA | 3 and site management plans to ensure the safe use of PFAS over the transition period. More information on the specific derogations can be found in Table 9 within the restriction proposal. 1.3 Structure of this report The remainder of this report is structured as follows: Section 2 - Scope and methodology: describes and explains the derogations and the variations made to them during the assessment as well as the approach and assumptions made throughout the assessment. Section 3 - Fgas maintenance derogation cost estimates: provides detailed explanation of the calculation of the cost variables considered during this assessment along with a population calculation and presentation of the legal scenario costs and benefits. In addition to these sections this report comes with several additional annexes containing supplementary data: Annex 1 - ICE vehicle sales estimate data: provides a detailed breakdown of the ICE vehicle sales data used by country and by year for the purpose of this assessment and outlines the decline model which has been applied to ICE sales during the transition to battery electric. Annex 2 - Detailed scenario cost breakdowns: contains a detailed breakdown of each legal scenarios cost by cost variable and by year of occurrence. Also includes a foreword advising on the nuanced way in which this detail should be interpreted and communicated. Annex 3 - Cost variable data tables: contains a detailed breakdown of each cost variable across each legal scenario in line with the calculations which are described and explained in section 3 of this report. MAC Maintenance Derogation Impact Assessment RPA | 4 2 Scope and methodology 2.1 Maintenance derogations assessed in this study The impacts on the derogations that are assessed in this study are in relation to mobile air conditioning (MAC) systems in internal combustion engine (ICE) vehicles. The current wording of the proposed derogations is not entirely clear. The current understanding of these derogations, which have formed the basis of the subsequent economic assessment by the RPA study team are found below. The two derogations within the restriction proposal relating to MAC systems in ICE vehicles are as follows: Paragraph 5: By way of derogation, paragraphs 1 and 2 shall not apply to (selection of relevant articles): (5.i) Maintenance and refilling of existing heating, ventilation, airconditioning, and refrigeration (HVACR) equipment put on the market before [18 months after entry into force (EiF)] and for which no dropin alternative exists until 13.5 years after EiF; and (5.p) Refrigerants in mobile air conditioning systems in combustion engine vehicles with mechanical compressors until 6.5 years after EiF. The RPA study team has made the inference that derogation 5.i - maintenance and refilling of existing HVACR equipment refers generally to HVACR equipment and includes MAC systems for ICE vehicles. In the Annex XV Report, the Dossier Submitters set some examples of HVACR: "domestic, commercial and industrial refrigeration, mobile and stationary air conditioning, and heat pumps", and they add that this derogation is on "all fluorinated gases used in commercial and industrial refrigeration, mobile and stationary air conditioning" (page 149). Therefore, we understand that this maintenance and refilling derogation would apply to MAC equipment. Whilst the other derogation 5.p - refrigerants in mobile air conditioning systems in combustion engine vehicles with mechanical compressors refers to ICE vehicles placed on the market using Fgases as the refrigerant. It is assumed that, by way of derogation 5.i, that maintenance and refilling of MAC systems in ICE vehicles can continue for 13.5 years after EiF and by way of derogation 5.p the sale of new ICE vehicles containing a fluorinated gas refrigerantbased MAC can continue for 6.5 years EiF. 2.1.1 Scenario variations The two above derogations individually present no obvious issue for the market from a compliance point of view. However once combined a slight legal discrepancy has been identified and this has led to this assessment developing and applying two scenarios for consideration. The issue is around the maintenance derogation 5.i stating that no HVACR system, including MAC systems, are sold after the initial 18month EiF derogation. However, derogation 5.p says that ICE vehicles containing a MAC system powered by an Fgas refrigerant, such as HFOs, can continue to be sold for the initial 18month derogation period plus an additional five years. There is therefore the possibility that vehicles sold during the 5year 5.p derogation period cannot have their MAC maintained to preserve its lifespan. RPA has had input from legal experts that this is the strictest interpretation, and most likely outcome, of the current restriction proposal, and this input has also highlighted there is a legal argument for this wording to be amended and softened to allow for vehicles sold in the 5.p derogation period to be included under the 5.i maintenance derogation. As a result of this RPA has conducted a cost MAC Maintenance Derogation Impact Assessment RPA | 5 assessment of the maintenance derogation based on two possible scenarios which are explained below. Scenario 1 This scenario assumes that the restriction proposal will enter into force at the start of January 2026. The initial 18month derogation period will therefore cease at the start of July 2027. From this point onwards under derogation 5.p ICE vehicles can be sold containing a MAC system powered by an Fgas until the beginning of July 2032 with maintenance of Fgas powered systems permitted until the start of July 2039 under derogation 5.i. The scenario 1 assessment however assumes that ICE vehicles containing an Fgas MAC sold between July 2027 and July 2032 are included under the 5.i maintenance derogation and can continue to be maintained and refilled until the start of July 2039. Scenario 1 is therefore an assessment of costs to the European economy based on a revision of the current wording of the PFAS restriction proposal. Scenario 2 This scenario assumes that the restriction proposal will enter into force at the start of January 2026. The initial 18month derogation period will therefore cease at the start of July 2027. From this point onwards under derogation 5.p ICE vehicles can be sold containing a MAC system powered by an Fgas until the beginning of July 2032 with maintenance of Fgas powered systems permitted until the start of July 2039 under derogation 5.i. In scenario 2 the assessment will be conducted according to the current interpretation of the restriction proposals wording. The scenario 2 assessment will be conducted based on the assumption that ICE vehicles sold between the start of July 2027 and end of June 2032 are not permitted to be maintained under derogation 5.i. 2.2 Aim of study The aim of this study is to estimate the costs and benefits to the European economy and consumers due to the 5.i maintenance derogation with regards to its impacts on ICE vehicles containing an Fgas powered MAC system. Due to the 5.i maintenance derogation not being indefinite there is a possibility that some European consumers will see early retirement of MAC equipment (i.e., the vehicle) they have purchased, and this will result in a cost to themselves, while also incurring a possible benefit from asset resale. Some consumers may have to scrap their vehicle entirely which could incur premature new vehicle purchase costs. The following section outline these general assumptions (applied to both scenarios outlined above) and the variables which have been considered and quantified to arrive at a final assessment of the impact caused by the 5.i maintenance derogation under both scenarios. 2.2.1 Assumptions Below is a list of the general assumptions the RPA study team has used in both scenario impact assessments. 1. An internal combustion engine (ICE) vehicle has a lifespan of 18 years5; 2. The lifespan of an EV is also assumed to be 18 years; 3. Mobile air conditioning (MAC) equipment has the same lifespan as an ICE vehicle when it can be maintained; 5 https://etrr.springeropen.com/articles/10.1186/s12544020004640 accessed July 2023 MAC Maintenance Derogation Impact Assessment RPA | 6 4. MAC equipment only lasts 6 years before needing to be refilled6 and therefore when it cannot be maintained will only have a lifespan of 6 years; 5. Based on 18year vehicle lifespan and 6year MAC charge lifespan ICE vehicles sold from July 2027 to the end of June 2032 are within scope of this assessment; 6. All consumers will perform one final recharge of their MAC system in 2039 to maximise the usable life of their MAC and vehicle, meaning vehicles sold from July 2027 onwards will see some early retirement of their vehicle (vehicles sold in 2027 will see 0.5 years of service life lost, 2028 vehicles 1.5 years and so on); 7. The final recharge in 2039, and 6year MAC charge lifespan assumptions result in retirement of vehicles and subsequent costs and benefits to be realised by consumers and the economy in 2045; 8. No new ICE vehicles containing an Fgas powered MAC can be put on the market from the start of July 2032 (It is expected that manufacturers will supply heater only models until 2035 due to unviable nature of developing new MAC systems for ICE vehicles before the ICE vehicle ban comes into force7); 9. European countries are divided into two climate groups. The first climate group will be a combination of European climates with very hot or very cold climates where not having a MAC poses safety concerns such as harmful internal car temperatures for both engines and passengers and defogging in cold climates, this will be referred to as the `scrappage' climates. The second climate group will be more temperate climates where consumers enjoy having a MAC system, but it is not connected to the vehicle or drivers' overall safety, this will be referred to as the `continuation' climates; 10. European countries which have a mix of both climates have been divided 50/50 into each of the climate groups; 11. After maintenance derogations end people in scrappage, climates will opt to prematurely scrap their vehicle and purchase a new one due to the safety concerns for passengers and the vehicle itself by not having active cooling or defogging from a MAC system; 12. After prematurely scrapping their vehicles consumers in scrappage climates will sell their vehicle second hand outside of Europe. It is assumed that export of second hand PFAS equipment regardless of PFAS concentration to outside the EU will continue to be permitted; 13. The resale of secondhand vehicles containing a depleted Fgas powered MAC system will violate internal market PFAS concentration limits under the restriction proposal and therefore cannot be resold within Europe; 14. Any prematurely scrapped vehicle with less than one full years' worth of service life remaining at the time of scrappage won't be resold second hand and instead will just be completely scrapped; 15. Following the sale of their vehicle, scrappage climate consumers would favour purchasing a new EV (most likely a battery electric vehicle, BEV) with a MAC system charged with a non fluorinated alternative; 16. People in continuation climates will see an early retirement of their MAC system (resulting in a loss of utility from a now inferior product) as it cannot be maintained any longer and cannot be replaced, but they will not scrap their entire vehicle; 17. MAC systems which are still operational following the end of the MAC derogation will be permitted to continue to operate beyond 2032 until the MAC system reaches the end of its final charge8; 6https://www.tesla.com/ownersmanual/model3/en_gb/GUIDE95DAAD9646E42499930 B109ED7B1D91.html accessed July 2023 7 RPA (2022): SocioEconomic Analysis and Impact Assessment of a potential REACH Restriction on Fgases as PFAS, report for Chemours Thermal Specialised Solutions (TSS), December 2022, Norwich, Norfolk, UK 8 https://echa.europa.eu/documents/10162/2156610/230405_upfas_webinar_qa_ds_en.pdf/3f47fdcc17c5 4b37b758720bb7e462f3?t=1687893645025 Question 2.3, accessed July 2023 MAC Maintenance Derogation Impact Assessment RPA | 7 18. The assumed lifespan of Fgas MAC system garage refilling equipment is assumed to be 10 years, therefore meaning any equipment purchased between the beginning of July 2030 and the end of June 2039 will see some early retirement; 19. ICE and battery electric vehicles (BEV) will reach price parity in 20269 which will be achieved by ICE vehicles increasing in price over time and BEV vehicles declining steadily in time, therefore all vehicles affected by service life lost will be assumed to have been sold at the same price calculated from a base year. Additional information on the use and justification for these assumptions is presented throughout section 3. After identifying the above 2 scenarios for assessment and the stated list of assumptions a methodology for assessment was devised and is explained in the following section. 2.2.2 Approach and methodology By use of the above general assumptions the RPA study team created a general formula which will serve as the basis for both scenario assessments. In total five variables were identified as quantifiable costs under both scenarios, two unquantifiable utilitybased costs were identified, and one quantifiable benefit was identified. The quantifiable costs that were identified are split across two different populations in line with the above assumptions relating to different European climate groups and the different assumed behaviours of consumers in those climates. These cost and benefit variables have been combined with these population parameters to deduce the following total cost assessment calculation for both scenarios outlined above. A description of the elements then follows: = ( 1) + ( 1) + ( 1) + ( 2) + + ( 2) + ( 1) - ( 1) n1: The number of cars sold in scrappage climates, consisting of very hot and cold climates, between July 2021 and the end of June 2032. This figure has been calculated per year with 2021 and 2022 values being informed by reported sales and post 2022 values being estimated using an estimated ICE sales decline model from previous work for TSS by RPA10. These vehicles will be completely scrapped and sold outside the EU following the end of the 5.i maintenance derogation. n2: The number of cars sold in continuation climates, consisting of more temperate climate zones, between July 2021 and the end of June 2032. This figure has been calculated per year with 2021 and 2022 values being informed by reported sales and post 2022 values being estimated using an estimated ICE sales decline model from previous work for TSS by RPA10. These vehicles will not be scrapped at the end of the 5.i derogation period and instead will continue operating without a functioning MAC system. ICE retirement value loss: This is the remaining discounted value of an ICE vehicle left at the end of June 2039 when the vehicle will be scrapped by consumers in scrappage climates. This value has been determined by dividing the 9https://www.autocar.co.uk/carnews/industrynewsenvironmentandenergy/richardparryjonescost paritybetweenevsandice accessed July 2023 10 RPA (2022): SocioEconomic Analysis and Impact Assessment of a potential REACH Restriction on Fgases as PFAS, report for Chemours Thermal Specialised Solutions (TSS), December 2022, Norwich, Norfolk, UK MAC Maintenance Derogation Impact Assessment RPA | 8 average cost of an ICE vehicle by 18 years to estimate the annual value with each annual value then being discounted at a 4% rate. Then, summing backwards from the 18th year the loss of value for a vehicle sold in a specific year was determined. For example, the ICE retirement loss of a vehicle scrapped in July 2039 which was purchased in July 2021 would be equal to half the 18th discounted year value of an average ICE vehicle (a 2022 vehicles loss value would be the entire 18th year value plus half the 17th year value and so on). This calculation is further expanded on in section 3. Premature EV cost incurred: Scrappage climate consumers will transition to an electric vehicle after scrapping their ICE vehicle. However, because these consumers have prematurely scrapped their ICE vehicle, they have subsequently had to prematurely purchase a new EV meaning this cost has been brought forward to the consumer ahead of when they would have originally planned. Therefore, these costs are a directly attributable impact caused by the occurrence of the 5.i derogation ending. This value has been determined by dividing the average cost of an EV by 18 years to estimate the annual value with each annual value then being discounted at a 4% rate. Then, summing forward this time from the 1st year of cost brought forward. For example, a consumer who purchased a vehicle in July 2021 which see half the first year of cost for their new EV brought forward prematurely (a 2022 vehicle would see the entire 1st year value plus half the 2nd year value and so on). The calculation of this variable is further expanded on in section 3. New EV production emissions: To produce an electric vehicle an amount of carbon dioxide is subsequently emitted during the manufacturing process. Therefore, due to additional electrical vehicles needing to be manufactured to fulfil the demand of scrappage climate consumers prematurely scrapping their vehicles, this level of additional carbon dioxide emissions is directly attributable to the 5.i derogation. An average amount of CO2 being required to manufacture an EV has been researched and this volume of carbon has subsequently been valued by multiplying this volume in tonnes by a value of 84/tonne11. MAC retirement loss: Consumers in continuation climates are assumed to not scrap their vehicle after July 2039 but will instead use their vehicle without a functioning MAC system. Therefore, the value for MAC retirement cost has been estimated by discounting the estimated value of a MAC system annually over an 18 year period. Then, summing backwards from the 18th year the loss of value for a MAC sold in a specific year was determined. For example, a MAC retired in July 2039 which was purchased in July 2021 would be equal to half the 18th discounted year value of MAC system (a 2022 MAC retirement loss value would be the entire 18th year value plus half the 17th year value and so on). The calculation of this variable is further expanded on in section 3. HFO equipment retirement costs: Local garages throughout Europe currently have equipment designed to conduct refilling of current F gas powered MAC systems. From July 2039 onwards this equipment will become obsolete and as such any equipment purchased from July 2030 onwards will see some level of early retirement. This value has been estimated by calculating the value of equipment discounted over a tenyear period according to how much usable life is left at the time of retirement. MAC retirement utility loss: A consumer in a continuation climate in Europe will have gained some utility from the use of their MAC during the use of their ICE vehicle, such as the removal of moisture from the passenger cabin of 11 State and Trends of Carbon Pricing 2022 https://openknowledge.worldbank.org/handle/10986/37455 The nominal price per tonne of carbon emitted is said to be $87 USD/tCO2e, the price was converted to euros on an exchange rate from Dollars to Euros of $1 = 0.97 MAC Maintenance Derogation Impact Assessment RPA | 9 the vehicle. The early retirement of this equipment will leave the consumer with a now inferior and worse performing product resulting in a loss of utility for the consumer. This variable, given its complexity, has not been quantified and is discussed qualitatively. New EV MAC system lower efficiency utility loss: Previous RPA consultation with MAC manufacturers and vehicle OEMs10 identified nonHFO powered MAC systems to be inferior in performance to their HFO powered counterparts. Consequently, consumers in scrappage climates who prematurely transition to an electric vehicle will see a loss in utility from a less efficient MAC system. In theory continuation climate consumers will also experience this loss of utility but because they will not have transitioned prematurely because of the 5.i derogation, the utility loss of continuation climate consumers is considered out of scope. This variable, given its complexity, has not been quantified and is discussed qualitatively. Value of exported retired car: This variable is the value of all the vehicles scrapped at the end of the 5.i derogation period in 2039. This value is deducted from the total cost figure as this value is returned to consumers and the European economy and is therefore a positive effective of the 5.i derogation. The value of an exported vehicle has been estimated to be 45% of the original vehicle price based on a 55% rate of depreciation in the secondhand market12. The value of an exported vehicle will not change between scenarios, and neither will the number of vehicles being exported as n1 will not change so the value of exported cars variable will remain constant across scenarios. The n1 variable for this calculation will be reduced by the number of vehicles with less than one full year of service life remaining at the time of retirement. 12 Car Depreciation UK Market Guide (2023 Update) | Motorway MAC Maintenance Derogation Impact Assessment RPA | 10 3 Fgas maintenance derogation cost estimates 3.1 Vehicle sale estimations (n1 and n2 estimates) The first step for the assessment of impacts caused by the 5.i maintenance derogation under the two proposed scenarios was to define the impacted population. As previously discussed, there are two populations which are impacted in different ways by the 5.i derogation within Europe. Consumers in very hot and cold climates referred to as scrappage climates and consumers in cool and temperate climates, referred to as continuation climates. This section will outline which European countries have been classed under each climate type followed by an explanation of how the n1 and n2 populations have been estimated. 3.1.1 Climate determinations It is noted that the most common classification system for climate types is the Koppen classification system, which consists of five main climate zones and multiple subzones within each main zone13. For this assessment several of these climate zones have been merged for simplicity to divide Europe into scrappage and continuation climate groups. In addition to this Europe is commonly divided into several accepted regions, Southern Europe, Western Europe, Central Europe, Northern Europe, Eastern Europe, and SouthEastern Europe. Figure 31 below presents a map of these accepted geographical groupings. Figure 31: Commonly accepted geographical areas of Europe Source: https://wonderingmaps.com/regionsofeurope/ This geographical distinction served as an initial basis for the assessment of which European nations may be classed under each climate type. Meteorological information and climate maps were then applied to this initial assessment to further refine which nations should be classed in each climate type. Figure 32 is the European Environment Agencies main climates of Europe map which was used for the final determination phase of the climate assessment. It should be noted that from the below 13 https://study.com/learn/lesson/europeclimatezonesclassifications.html MAC Maintenance Derogation Impact Assessment RPA | 11 diagram climates 14 and 813 were deemed by the study team as scrappage climates with climates 57 being classed as continuation climates. Figure 32: Main climates of Europe Source: https://www.eea.europa.eu/dataandmaps/figures/climate After using this meteorological information, a near complete list of European countries under each climate type was drawn up. However, the issue of mixed climates is then addressed for the final list. In total five European countries were identified as having a mixed climate, where part of the country is covered by a scrappage scored climate (14 or 813) and others a continuation scored climate (57). These five countries were Germany, Sweden, Slovenia, Croatia, and Bulgaria. Using Figure 32 above Germany and Sweden were determined by the study team to classify as a mixed climate country. MAC Maintenance Derogation Impact Assessment RPA | 12 Slovenia, Croatia, and Bulgaria required closer examination for a final determination and the below Figure 33 is the map used for the final assessment by the RPA team. A B C Figure 33: Slovenia (A), Croatia (B) and Bulgaria (C) climate map Source: https://www.eea.europa.eu/dataandmaps/figures/climate It was finally determined that more of Slovenia was classed as temperate or cold so this nation would be classed under that grouping, but Croatia and Bulgaria were seen to be sufficiently split in climate to be classed as mixed climates like France. Below in Table 31 is the final list of European countries assigned to each climate type. Table 31: Climates of European nations Continuation climates Mixed climates Scrappage climates Austria Bulgaria Belgium Czechia Croatia Cyprus Estonia Germany Denmark Hungary Sweden Finland Latvia France Lithuania Greece Poland Ireland Romania Italy Slovakia Luxembourg Slovenia Malta* Netherlands Portugal Spain *Malta has been classified as a scrappage climate in this assessment, but no vehicle sales information for Malta could be obtained. Therefore, Malta has been omitted from this assessment entirely. MAC Maintenance Derogation Impact Assessment RPA | 13 3.1.2 n1 and n2 vehicle sales With European countries now sorted into their respective climate groups it was possible for the RPA team to estimate the number of vehicles sold per country between 2027 and 2032. Sales information for the number of passenger cars sold per country for 2021 and 2022 was obtained14 and to this an extrapolation was made from 2023 to 2032. This extrapolation was taken from previous vehicle sales estimate information produced by RPA for Chemours TSS and followed the same declining rate estimation for the number of ICE vehicles being sold per year15. The rate of sales of ICE vehicles are assumed to fall over the period due to the structural transition over to the electrical vehicles prior to the implementation of an ICE vehicle sales ban in 2035. Just over 9.7 million ICE passenger vehicles were sold in 2021 with this expected to decline steadily down to around 2.48 million by 2032. Please note 2027 and 2032 figures in the graph below are halved due to only half of each year being within the assessment period. The annual rate of decline in ICE sales is estimated be between 11.8% and 12.2% per annum. Figure 34 graphically depicts the total number of vehicles sold per annum within Europe (excluding Malta) and the estimated rate of decline per annum. A table of exact sales figures, decline rate percentages and subsequent multipliers can be found in Annex 1 under section A1.2. Number of ICE vehicles sold Annual rate of decline in ICE sales Annual volume and rate of decline of ICE vehicles sold in Europe 5,000,000 4,500,000 4,000,000 3,500,000 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 0 2027 2028 2029 2030 Year 2031 2032 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Vehicles sold Decline rate (%) Figure 34: Annual volume and rate of decline of ICE vehicles sold in Europe Source: Refer to footnotes 14 and 15 for the sources of this information A detailed breakdown of the above total sales figures by European country according to their climate classification can be found in Annex 1 under section A1.1. Between July 2027 and June 2032, it is estimated that just over 2.44 million ICE passenger vehicles will be sold in continuation climates, with just over 5.86 million vehicles sold in mixed climates and over 9.72 million units estimated to be sold in scrappage climates over the period. To calculate the n1 variable for the impact assessment the study team per year summed the sales the of scrappage climates with 50% of the annual sales from mixed climate countries. Then to calculate the n2 population variable the values of continuation climates were summed with the remaining 50% of mixed climate sales. Table 32 below presents the total 14 https://www.bestsellingcars.com/europeannewcarsalesstatisticslinks/ 15 RPA (2022): SocioEconomic Analysis and Impact Assessment of a potential REACH Restriction on Fgases as PFAS, report for Chemours Thermal Specialised Solutions (TSS), December 2022, Norwich, Norfolk, UK MAC Maintenance Derogation Impact Assessment RPA | 14 annual sales figures for each climate group representing the two population variables necessary for the impact assessment equation outlined in section 2.2.2 to be calculated. Table 32: Annual vehicle sales per climate grouping representing the n1 and n2 population variables Year Scrappage climate vehicles (n1)** Continuation climate vehicles (n2) 2027* 1,710,000 728,000 2028 3,010,000 1,280,000 2029 2,650,000 1,130,000 2030 2,330,000 992,000 2031 2,050,000 871,000 2032* 903,000 384,000 Total 12,653,000 5,385,000 *2027 and 2032 represent sales figures for only half the year. **Excluding Malta 3.2 Scenario 1 cost estimate This section will provide a more detailed explanation of how each variable and element of the cost equation in section 2.2.2 was calculated. It will also subsequently explain further how each variable was used to calculate an estimate for the total cost to European consumers and society. Then the results to the scenario 1, as explained in section 2.1.1, cost assessment of the 5.i maintenance derogation are presented and discussed. 3.2.1 Cost variable calculations The first element of the total cost equation calculated by the RPA team was the ICE retirement value loss. However due to the ICE and BEV price parity assumption as stated in section 2.2.1 under point 19 results in the calculation of premature EV costs being incurred calculated in tandem. First the average price of a passenger vehicle in the EU27 and the UK was obtained for 2015 and 202016. Then via use of linear extrapolation accounting for the mean annual price increase the 2020 average passenger car price of 32,035 was inflated by a rate of about 2.53% to arrive at a 2021 value of 32,845. At the same time the average price for an electric vehicle in the EU in 2021 of $48,000, or 46,560 (based on a USD:EUR exchange rate of 1:0.97), was obtained from a study conducted by the International Energy Agency17. In line with assumption 19 these prices will reach parity by 2026 due to ICE vehicles becoming more expensive and BEVs becoming cheaper. To determine this price parity, point the possible fall in the price of a BEV was estimated. The best way of estimating the future price point of a BEV is to estimate the changes in lithiumion batteries; this is because around 51% of an electric vehicles cost is determined by the power train where its' batteries are housed18. Therefore, meaning that 23,745.60 of a BEVs' value comes from the power train. Between 2021 and 2025 battery costs have been estimated to fall by between 40% and 60% by several large vehicle manufacturers18 and so for the purpose of this assessment it has been assumed that by 2025 battery prices will have reduced by 50%. As a result, by 2025 the price of a BEVs' power train, and subsequently the price of a BEV overall, is estimated to have reduced in price 16 https://www.statista.com/statistics/425095/eucarsalesaveragepricesinbycountry/ accessed July 2023 17 https://iea.blob.core.windows.net/assets/ad8fb04c4f7542fc973a 6e54c8a4449a/GlobalElectricVehicleOutlook2022.pdf Page 27, accessed July 2023 18https://www.treehugger.com/willevcostsgodown accessed July 2023 MAC Maintenance Derogation Impact Assessment RPA | 15 by 11,872.80. After deducting this price reduction amount from the original BEV cost it is estimated and assumed that from 2025 onwards both ICE and BEVs will be sold at an average price of 34,687. Assumption 4 above states that a MAC system only lasts 6 years before needing to be refilled and consequently assumption 6 states that following a final recharge in 2039 only vehicles sold from July 2027 onwards will see some early retirement. Therefore, the average vehicles sales price of 34,687 for ICE and BEV has been set as the assumed sales price for all vehicles across the period. After establishing an average vehicle price for 2027 this value was then divided by the average lifespan of a vehicle of 18 years resulting in an annual price to the consumer of about 1,927. Then each annual value was discounted at a rate of 4% to estimate the annual present value of an ICE vehicle with 2027 as the base year (for example for 2028: 1,927*0.96=1,853, for 2029: 1,927*0.92=1,782 and so on). The ICE scrappage cost for a vehicle purchased in the second half of 2027 was then calculated to be half the 18th discounted year value of an ICE vehicle, similarly for a vehicle purchased in 2028 the scrappage cost is equal to the whole 18th discounted year plus half the 17th discounted year and so on until 2032. In monetary terms this means a vehicle purchased after July 2027 when scrapped in 2045 results in a loss to the consumer of about 495 but a vehicle purchased in 2032 and scrapped in 2045 results in a loss of 5,960 worth of value. Then for each year of the assessment period the annual scrappage value loss per vehicle was multiplied by each year's corresponding n1 value. The result of this calculation is presented and discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 7 under section A3.1. Similarly, to calculate the premature EV cost incurred value the estimated cost of an EV in Europe (also set at 34,687) has also been divided by 18 years to arrive at the annual cost. Similarly, to the above ICE retirement value loss calculation each annual value was then discounted at a rate of 4% per annum. The difference this time however is because we are assessing the cost of an EV being brought forward prematurely the cost, therefore, to consumers was calculated by summing each discounted annual value going forward instead of backwards. A consumer who purchased an ICE vehicle in 2027 will see half the first year's cost of an EV brought forward prematurely (estimated to be about 964) whereas a consumer who purchased an ICE vehicle in 2032 will see the first 5.5 years (estimated to be 9,714) of EV cost brought forward to them prematurely when they scrap their vehicle in 2045. Then for each year of the assessment period the annual premature cost per EV has then been multiplied by that year's corresponding n1 value. The result of this calculation is presented and discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 8 under section A3.1. Next the value of CO2 emissions from the production of these additional required EVs was quantified. This is not a direct cost to consumers unlike the previous two variables but is instead a wider cost to European society. As stated under section 2.2.2 the World Bank has estimated the cost of a tonne of carbon dioxide emissions to be worth around $87 in the EU which has been converted to a value of 84/tonne11. With the value of carbon dioxide emissions identified the next requirement was an estimate as to the volume of CO2 emitted during the manufacture of an electrical vehicle. Research conducted by The Swedish Environment Institute looking into the life cycle impacts of lithiumion batteries, including the kind used in BEVs, identified that an on average 17.5 tonnes of carbon dioxide are released during the manufacturing of an electric vehicle19. It is not clear if this value refers only to the CO2 emissions to produce the lithium batteries of an EV or if it is inclusive of the carbon emissions for other parts of the vehicle such as the steel body and electrical circuitry. This volume of emissions per vehicle was then multiplied by the number of additional EVs required by year with this total volume then being multiplied by the price per tonne of 84. The result of this calculation is presented 19 https://eartheclipse.com/environment/largeco2emissionsbatterieselectriccars.html accessed July 2023 MAC Maintenance Derogation Impact Assessment RPA | 16 and discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 13 under section A3.3. MAC system retirement losses were then calculated, and this calculation was conducted in the same way as the ICE retirement value loss variable described above. A MAC system was found to cost in the region of $1,500 with some models costing considerably more20, for this assessment however the lower end value of $1,500 was used which was converted on a 1:0.97 exchange rate to Euros in line with previous work with TSS by RPA. A MAC system is therefore estimated to cost 1,455. This price of MAC system was then divided across 18 years with each annual value being discounted at a rate of 4% per annum. The MAC retirement losses for a vehicle purchased in the second half of 2027 was then calculated to be half the 18th discounted year value of a MAC system, similarly for a vehicle purchased in 2028 the cost is equal to the whole 18th discounted year plus half the 17th discounted year and so on until 2032. In monetary terms this means a MAC purchased in 2027 when scrapped in 2045 results in a loss to the consumer of about 21 but a system purchased in 2032 and scrapped in 2045 results in a loss of around 250 worth of value. The only difference in this calculation is that the annual losses attributable to early MAC retirement were multiplied instead by the n2 population value as described in section 2.2.2. The result of this calculation is presented in discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 9 under section A3.1. Also, as previously mentioned utility losses from early retirement of MAC equipment and less efficient CO2 powered MAC systems in new EVs has not been estimated in this assessment. To calculate the cost of retirement of HFO refilling equipment an asset lifespan of 10 years was assumed as an exact lifespan could not be identified. By consulting several sources HFO refilling equipment was estimated to cost on average 2,87721 which was converted to a value of 3,424 based on an exchange rate of 1:1.19. Like the calculations above this value was then divided across the equipment lifespan with each year after year 1 being discounted at a rate of 4% per annum to estimate the scrappage cost per machine per year and like the ICE retirement value loss these were summed starting from the final year and working backwards. There are estimated to be around 380,000 local garages within Europe7 and it has been assumed that each local garage will have at least one piece of HFO refilling equipment. Specific sales of this equipment year on year could not be obtained during research to estimate annual sales of equipment between 2030 and 2039. In the absence of this specific estimate an assumed even sales distribution has been applied across the 10year period for the equipment resulting in an assumed 38,000 units being sold annually. This variable is therefore less robust than the other variables calculated in this assessment and carries a higher margin for over or underestimation. However as seen in the following sections the cost of early retirement of this equipment is substantially smaller than other costs considered in this assessment and as such any variation in the elements of this calculation will not substantially impact the final cost calculation. The result of this calculation is presented and discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 14 under section A3.3. The benefit value of the value generated from the sale of vehicles prematurely scrapped by consumers in scrappage climates was the final variable to calculate. To do this the average value of a passenger vehicle in Europe, as explained above to be estimated at 34,687, was reduced by 55% as explained in section 2.2.2. Therefore, meaning that it is estimated that a vehicle exported outside Europe and sold second hand will achieve a sale value of about 15,609. This value was then discounted at a rate of 4% per annum until 2045. However, unlike before where each individual discounted annual value was considered the only value used in calculation was the discounted value of the second vehicle in 20 https://cars.costhelper.com/carairconditioning.html accessed July 2023 21https://automotechservices.co.uk/products/asac2000airconservicemachiner1234yf/ and https://www.autoworkshopequipment.co.uk/Air_Con_Recharging_Machine.php and https://www.garageequipmentonline.com/products/airconditioningstations/ all accessed July 2023 MAC Maintenance Derogation Impact Assessment RPA | 17 2045 which was estimated to be 6,089. Only the 2045 value was used because the vehicles of consumers in scrappage climates won't scrap their vehicle until 2045 and as such will gain that year's value for their vehicle and not the value of the year, they purchased the vehicle. After estimating this value, it was then multiplied by the total value of the n1 population, minus the number of vehicles with less than one full years' service life left at retirement in line with assumption 14 (making the value used 10,944,753), to estimate the value returned to European consumers and society from the sale of their now obsolete vehicles. The result of this calculation is presented and discussed in the following section and a detailed, unrounded breakdown of this calculation can be found in Annex Table 15 under section A3.3. 3.2.2 Scenario results After conducting the calculations as described above using the above stated information the total cost calculation as stated in section 2.2.2 was conducted for scenario 1. The results of this can be found below in Table 33. Table 33: Scenario 1 calculation results Scenario variable Value () Early ICE scrappage cost 35,800,000,000 Premature EV cost 63,000,000,000 Early MAC retirement cost 639,000,000 Vehicle export earnings 66,600,000,000 EV CO costs 18,600,000,000 HFO equipment scrappage cost 513,000,000 Sum of costs* 118,552,000,000 Sum of benefits** 66,600,000,000 Scenario total cost*** 51,900,000,000 *This is the sum of Early ICE scrappage cost, premature EV cost, early MAC retirement cost, EV CO2 costs and HFO equipment scrappage cost. **This is equal to the vehicle export earnings. ***Calculated according to cost equation outlined in section 2.2.2 As can be seen from the above table, scenario 1 is estimated to incur a total of over 118.5 billion in costs for European consumers and European society. Over half of this cost (63 billion) is attributed to the premature cost of purchasing an EV being brought forward onto European consumers in scrappage climates. This is an incredibly high level of cost to force onto consumers, but this value is on top of an additional 35.8 billion worth of equipment value lost being suffered by those very same consumers. This is partially offset however with the sale of these scrapped vehicles estimated to return about 66.6 billion in value. Consumers in continuation climates are less impacted compared to their warmer climate counterparts with consumers in continuation climates suffering 639 million worth of equipment value loss in 2045. Wider European society also burdens a heavy price with local garages, most of which are SMEs, facing over half a billion Euros in scrappage costs and European society incurs an environmental cost of around 18.6 billion worth of additional CO2 emissions because of additional EVs needing manufacturing. Therefore, under a scenario where ICE vehicles containing a MAC powered by an Fgas sold after July 2027 are permitted to be maintained under the proposed 5.i maintenance derogation, European consumers and European society is estimated to incur a total cost of 51.9 billion. It should also again be noted that this value does not account for losses of utility, which will be suffered by European consumers from losing their MAC systems or having to use less efficient ones in new EVs. While there are some benefits generated from this scenario and by the maintenance derogation, they are MAC Maintenance Derogation Impact Assessment RPA | 18 categorically outweighed by the substantial costs that will also be incurred because of the regulatory action. 3.3 Scenario 2 cost estimate This section will provide a detailed explanation of the variations in how certain cost variables of the cost equation in section 2.2.2 were calculated. It will also subsequently explain further how each variable was used to calculate an estimate for the total cost to European consumers and society. Then the results to the scenario 2, as explained in section 2.1.1, cost assessment of the 5.i maintenance derogation are presented and discussed. 3.3.1 Cost variable calculations For the calculation of the scenario 2 total cost estimate the method for calculating vehicle export earnings, new EV production emissions and HFO equipment scrappage costs remains the same as described above. Therefore, the results for these three variables when presented in the following section will be the same value as those presented above in Table 33. The method employed under scenario 2 for calculating ICE retirement value loss, premature EV cost incurred, and MAC retirement cost is also very similar to that described above. However, under scenario 2 there is now a distinction for vehicles sold from July 2027 to the end of June 2032 as scenario 2 is based on the current interpretation of the 5.i derogation where these vehicles are not permitted to be maintained from the start of July 2027. The primary difference this causes in the calculation method is the years of service left at the retirement of the ICE vehicle which will impact the scrappage cost per vehicle. Research identified that a MAC system on average lasts around 6 years before needing to be refilled6. Therefore, meaning that after 6 years consumers in scrappage climates will scrap their vehicle as their MAC system will no longer be functional resulting in their vehicle having 12 years of service life left in it at the time of retirement. Likewise for consumers in continuation climates their MAC system will cease to work after 6 years leaving 12 years of functional life of the equipment unutilised. All vehicles within scope of this assessment are those sold after July 2027 therefore meaning the full n1 and n2 populations will suffer 12 years' service life lost. The scrappages cost of an ICE vehicle for scrappage climate consumers is estimated to be 14,865 with 18,809 worth of premature EV cost being brought forward to them. Continuation climate consumers will see a scrappage value loss of about 789 for their MAC systems. The result of these calculations are presented and discussed in the following section and a detailed, unrounded breakdown of these calculations can be found in Annex 3 under section A3.2. 3.3.2 Scenario results After conducting the calculations as described above using the above stated information the total cost calculation as stated in section 2.2.2 was conducted for scenario 2. The results of this can be found below in Table 34. Table 34: Scenario 2 calculation results Scenario variable Early ICE scrappage cost Premature EV cost Early MAC retirement cost Vehicle export earnings EV CO costs HFO equipment scrappage cost Value () 188,000,000,000 238,000,000,000 4,250,000,000 66,600,000,000 18,600,000,000 513,000,000 MAC Maintenance Derogation Impact Assessment RPA | 19 Table 34: Scenario 2 calculation results Scenario variable Value () Sum of costs* 449,363,000,000 Sum of benefits** 66,600,000,000 Scenario total cost*** 383,000,000,000 *This is the sum of Early ICE scrappage cost, premature EV cost, early MAC retirement cost, EV CO2 costs and HFO equipment scrappage cost. **This is equal to the vehicle export earnings. ***Calculated according to cost equation outlined in section 2.2.2 As can be seen from the above table, scenario 2 is estimated to incur substantially higher costs than scenario 1 with a total of over 449 billion in costs for European consumers and society. Again, over half of this cost (238 billion) is attributed to the premature cost of purchasing an EV being brought forward onto European consumers in scrappage climates. This is an incredibly high level of cost to force onto consumers, but this is not the only cost which is greater in scenario 2 with ICE vehicle scrappage costs also being higher at 188 billion. While these costs have increased however the value of early retired vehicles being resold remains unchanged at 66.6 billion in value. Consumers in continuation climates are again significantly less impacted compared to their scrappage climate counterparts but still see a substantial increase in the costs they will face with their estimated costs rising to 4.25 billion worth of equipment value loss in 2045. The costs suffered by wider European society from local garages scrapping their HFO equipment prematurely and the cost of additional carbon emissions remain unchanged under scenario 2 at 513 million and 18.6 billion respectively. Therefore, under a scenario where ICE vehicles containing a MAC, powered by an Fgas sold after July 2027 are not permitted to be maintained under the proposed 5.i maintenance derogation, European consumers and European society is estimated to incur a total cost of 383 billion. It should also again be noted that this value does not account for losses of utility, which will be suffered by European consumers from losing their MAC systems or having to use less efficient ones in new EVs. While there are some benefits generated from this scenario and by the maintenance derogation, they are categorically outweighed by the substantial costs that will also be incurred because of the regulatory action. 3.4 Conclusions and recommendations The derogations proposed in the PFAS restriction dossier are designed to limit impacts and damages to the European economy, consumers, and society by aiding parties with a transition time to orderly move to alternative technologies and materials. It is clear however that the proposed 5.i maintenance derogation in the restriction proposal will result in some very significant, and damaging costs to European consumers and society. This assessment has looked at two possible scenarios for the currently proposed 5.i derogation and both scenarios result in billions of Euros in costs for consumers and society being inflicted in 2045. The results of both scenarios cost calculations are presented side by side below in Table 35 for comparison. Table 35: Scenario 1 and 2 calculation results and comparison Scenario variable Scenario 1 value (, billions) Early ICE scrappage cost 35.8 Premature EV cost 63 Early MAC retirement cost 0.64 Vehicle export earnings 66.6 EV CO costs 18.6 HFO equipment scrappage cost 0.513 Scenario total cost* 51.9 Scenario 2 value (, billions) 188 238 4.25 66.6 18.6 0.513 383 MAC Maintenance Derogation Impact Assessment RPA | 20 Table 35: Scenario 1 and 2 calculation results and comparison Scenario variable Scenario 1 value (, billions) * Calculated according to cost equation outlined in section 2.2.2 Scenario 2 value (, billions) Scenario 2 is based on the currently understood interpretation of the 5.i maintenance derogation, making the scenario 2 results the currently most likely outcome for Europe to experience. At present it is unclear if this is the exact intention of the dossier submitters, but the current wording and interpretation of the derogation makes scenario 2 most likely. In the event the scenario 2 legal interpretation was unintentional then scenario 1 would be most likely to occur. Therefore, meaning it is estimated that the 5.i maintenance derogation as currently framed will result in around 383 billion of cost being imposed upon European consumers and society in 2045. It should again be noted that these costs while calculated based on vehicles sales between July 2027 and the end of June 2032, the cost of this derogation will be realised in 2045. The positive value of 66.6 billion from early retired ICE vehicles being exported outside the EU is significantly outweighed by the costs imposed by this regulatory action. It should also be noted that while this assessment considers the value from export vehicles as a benefit, in real terms if second hand vehicles containing an Fgas powered MAC system are, exported outside the EU where their MAC systems can be refilled the 5.i maintenance derogation will not result in a reduction in Fgas emissions on a global scale. This is because with a glut in supply now present in third party countries, consumers who previously could not afford a car now might be able to on top of the base demand for motor vehicles. Therefore, additional motorists globally could begin using ICE vehicles containing refilled Fgas MAC systems which would continue to output additional PFAS emissions. This could therefore support an argument for the prevention of exports of this retired equipment which is currently under consideration by regulators. However, imposing such a ban would be even more harmful for consumers. This is because while an export ban would prevent the subsequent refilling of HFO MAC systems in third party countries and resultant emissions from them, an export ban would lead to an additional 66.6 billion in cost being imposed on European consumers. Not only that however but if an export ban were to be implemented there would be considerable scrappage logistical issues with Europe not having sufficient scrappage and storage capacity to deal with millions of vehicles being scrapped within a short period. An export ban therefore on environmental, logistical, and economical grounds could be considered wholly impractical and counterproductive. The results of scenario 2 therefore clearly indicate that the current application of the 5.i derogation will impose exceptionally damaging costs onto consumers and society. Scenario 1 which constitutes a slight softening of the current application of the 5.i derogation was therefore considered as a means of a possible cost limitation strategy. Adapting the current application of the derogation as outlined in scenario 1 results in a significant 85% reduction in costs to European consumers and society. However, scenario 1 still results in 51.9 billion worth of costs being imposed on European consumers and society, with this rising by, again, an additional 66.6 billion if exports of retired equipment were to be prohibited as well. Therefore, while a softening of the scenario around the 5.i maintenance derogation is possible and does substantially reduce the costs estimated to be incurred by consumers and society the costs remain high and would be damaging to European society but even more so to consumers. The costs and results presented in Table 35 are also presented graphically below in Figure 35. MAC Maintenance Derogation Impact Assessment RPA | 21 Value (, billions) Economic cost comparison of legal scenarios to the proposed Fgas maintenance derrogation 450 400 350 300 238 250 188 200 150 100 35.8 63 50 0 50 100 Early ICE Premature EV scrappage cost cost 4.25 0.64 Early MAC retirement cost 383 18.6 0.513 51.9 18.6 0.513 66.6 66.6 Vehicle export earnings Cost Variable EV CO costs HFO equipment scrappage cost Scenario total cost* Scenario 1 Scenario 2 Figure 35: Economic cost comparison of the two scenarios to the proposed 5.i maintenance derogation In conclusion the proposed 5.i derogation will result in significant costs being imposed on consumers and society in 2045. The current permittance of exports for retired equipment allow costs under both scenarios assessed to be partially mitigated, although exports will not see the global levels of Fgas emissions from MAC reduce. It is illogical however to prevent exports to prevent further Fgas emissions as this results in an intolerable level of additional cost being levied against consumers and society. Cost mitigation is possible under the current proposed derogation with minor adaptations to the legal wording and application, but while reducing costs substantially, large costs continue to be imposed on consumers under an amended legal wording. Therefore, the conclusion of this assessment is that any time limited maintenance derogation will result in an excessive level of cost being endured by consumers and society. A suitable recommendation would be for the 5.i derogation to be amended to allow maintenance of Fgas equipment indefinitely. This would prevent undue costs being imposed on consumers and society entirely and it would still be in line with the desire to reduce Fgas emissions. This because while maintenance would be allowed to continue indefinitely, equipment would eventually fail and be replace with nonPFAS alternatives which would still result in a fall in Fgas emissions over the long term, but this would no longer come at a prohibitive cost to the economy, consumers, and society. MAC Maintenance Derogation Impact Assessment RPA | 22 Annex 1 ICE vehicle sales estimates data A1.1 European ICE vehicle sales estimates by country and climate Annex Table 1: Annual ICE vehicle sales estimates in European countries with a continuation climate Country 2027* 2028 2029 2030 2031 2032* Austria 56,700 99,700 87,800 77,200 67,800 29,900 Czechia 50,600 89,100 78,400 69,000 60,600 26,700 Estonia 5,690 10,000 8,800 7,750 6,800 3,000 Hungary 29,400 51,700 45,500 40,000 35,200 15,500 Latvia 4,410 7,750 6,820 6,000 5,270 2,320 Lithuania 6,730 11,800 10,400 9,170 8,060 3,550 Poland 111,000 195,000 171,000 151,000 132,000 58,400 Romania 34,100 60,000 52,800 46,400 40,800 18,000 Slovakia 20,800 36,600 32,200 28,300 24,900 11,000 Slovenia 12,200 21,500 18,900 16,600 14,600 6,450 Source: https://www.bestsellingcars.com/europeannewcarsalesstatisticslinks/ , ACEA and RPA analysis *Represents 50% of full year estimate Annex Table 2: Annual ICE vehicle sales estimates in European countries with a mixed climate Country 2027* 2028 2029 2030 2031 2032* Bulgaria 7,560 13,300 11,700 10,300 9,050 3,990 Croatia 11,300 19,900 17,500 15,400 13,500 5,970 Germany 699,000 1,230,000 1,080,000 952,000 836,000 369,000 Sweden 76,000 134,000 118,000 103,000 90,900 40,100 Source: https://www.bestsellingcars.com/europeannewcarsalesstatisticslinks/ , ACEA and RPA analysis *Represents 50% of full year estimate Annex Table 3: Annual ICE vehicle sales estimates in European countries with a scrappage climate Country 2027* 2028 2029 2030 2031 2032* Belgium 96,600 170,000 150,000 132,000 116,000 50,900 Cyprus 3,070 5,390 4,750 4,180 3,670 1,620 Denmark 39,100 68,800 60,500 53,300 46,800 20,600 Finland 21,500 37,900 33,300 29,300 25,800 11,400 France 403,000 709,000 624,000 549,000 482,000 213,000 Greece 27,800 48,800 43,000 37,800 33,200 14,600 Ireland 27,700 48,800 43,000 37,800 33,200 14,600 Italy 347,000 610,000 537,000 473,000 415,000 183,000 Luxembourg 11,100 19,500 17,200 15,100 13,300 5,850 Netherlands 82,300 145,000 127,000 112,000 98,500 43,400 Portugal 41,200 72,500 63,800 56,100 49,300 21,700 Spain 214,000 377,000 332,000 292,000 257,000 113,000 Source: https://www.bestsellingcars.com/europeannewcarsalesstatisticslinks/ , ACEA and RPA analysis *Represents 50% of full year estimate MAC Maintenance Derogation Impact Assessment RPA | 23 A1.2 ICE sales decline model data Annex Table 4: ICE vehicle sale estimates in Europe per year and estimated annual rate of decline in sales. Previous year multiplier Year Vehicle sales*** Decline rate (Dr) % (1Dr) % 2021* 9,700,000 0.00% 0.00% 2022* 9,260,000 0.00% 0.00% 2023 8,160,000 11.82% 88.18% 2024 7,180,000 12.06% 87.94% 2025 6,310,000 12.08% 87.92% 2026 5,550,000 12.00% 88.00% 2027** 4,880,000 12.12% 87.88% 2028 4,290,000 12.07% 87.93% 2029 3,780,000 11.96% 88.04% 2030 3,320,000 12.03% 87.97% 2031 2,920,000 12.15% 87.85% 2032** 1,290,000 11.82% 88.18% Source: https://www.bestsellingcars.com/europeannewcarsalesstatisticslinks/ and RPA (2022): Socio Economic Analysis and Impact Assessment of a potential REACH Restriction on Fgases as PFAS, report for Chemours Thermal Specialised Solutions (TSS), December 2022, Norwich, Norfolk, UK *Data based on actual sales and as such decline rate and multiplier have been set at 0 **Represents 50% of full year estimate ***Excluding sales in Malta. MAC Maintenance Derogation Impact Assessment RPA | 24 Annex 2 Detailed scenario cost breakdowns A2.1 Foreword The tables presented in this annex are provided for the purpose of seeing a detailed breakdown of the costs incurred and benefits enjoyed by the European economy and consumers from the two assessed scenarios of the proposed maintenance derogation. The tables provide a breakdown of the costs and benefits by each element as presented earlier in the report, but these breakdowns also provide a breakdown of the cost per year. However, it is necessary to provide some context to the presentation of this information to ensure accurate interpretation of the results these tables depict. The values in both tables are represented as occurring in a certain year (for example 1.69 billion of cost occurring in 2027) however this is not correct. All the costs and benefits depicted below are realised by the European economy and consumers at the beginning of the second half of 2045. Therefore, the tables below should be interpreted as representing each year of the assessment periods contribution to the total cost which is realised in 2045. That is to say that the values presented for 2027 for example are the costs and benefits generated by the 2027 ICE vehicles sales and so on. A2.1.1 Scenario 1 breakdown Annex Table 5: Scenario 1 total cost breakdown by year and element Early ICE scrappage Year cost () Premature EV cost () 2027 847,000,000 1,650,000,000 2028 4,530,000,000 8,590,000,000 2029 6,770,000,000 12,400,000,000 2030 8,500,000,000 15,000,000,000 2031 9,790,000,000 16,600,000,000 2032 5,380,000,000 8,770,000,000 2033 2034 2035 2036 2037 2038 2039 Total* 35,800,000,000 63,000,000,000 Early MAC retirement cost () 15,100,000 80,800,000 121,000,000 152,000,000 175,000,000 96,100,000 639,000,000 Vehicle export earnings ()** 18,300,000,000 16,100,000,000 14,200,000,000 12,500,000,000 5,500,000,000 66,600,000,000 EV CO costs () 2,520,000,000 4,430,000,000 3,900,000,000 3,430,000,000 3,010,000,000 1,330,000,000 18,600,000,000 HFO equipment scrappage cost () 4,570,000 13,900,000 23,600,000 33,700,000 44,200,000 55,100,000 66,400,000 78,200,000 90,500,000 103,000,000 513,000,000 MAC Maintenance Derogation Impact Assessment RPA | 25 Annex Table 5: Scenario 1 total cost breakdown by year and element Early ICE scrappage Early MAC retirement Vehicle export HFO equipment Year cost () Premature EV cost () cost () earnings ()** EV CO costs () scrappage cost () Source: RPA analysis *The total row is a rounded value of the precise cost sum and is not a sum of the rounded values in the rows above. There is therefore a possibility that a columns values may not exactly sum to the value in the total cell, the total cell is however accurate, and this is only a minor rounding discrepancy with no impact on accuracy or validity. **Note that in the cost calculation exported vehicle earnings are subtracted from the total cost (reducing the total cost) so the total scenario cost is not a sum of the bottom row totals. A2.1.2 Scenario 2 breakdown Annex Table 6: Scenario 2 total cost breakdown by year and element Early ICE scrappage Early MAC retirement Vehicle export HFO equipment Year cost () Premature EV cost () cost () earnings ()** EV CO costs () scrappage cost () 2027 25,400,000,000 32,200,000,000 575,000,000 2,520,000,000 2028 44,800,000,000 56,600,000,000 1,010,000,000 18,300,000,000 4,430,000,000 2029 39,400,000,000 49,900,000,000 890,000,000 16,100,000,000 3,900,000,000 2030 34,700,000,000 43,900,000,000 783,000,000 14,200,000,000 3,430,000,000 4,570,000 2031 30,500,000,000 38,500,000,000 687,000,000 12,500,000,000 3,010,000,000 13,900,000 2032 13,400,000,000 17,000,000,000 303,000,000 5,500,000,000 1,330,000,000 23,600,000 2033 33,700,000 2034 44,200,000 2035 55,100,000 2036 66,400,000 2037 78,200,000 2038 90,500,000 2039 103,000,000 Total* 188,000,000,000 238,000,000,000 4,250,000,000 66,600,000,000 18,600,000,000 513,000,000 Source: RPA analysis *The total row is a rounded value of the precise cost sum and is not a sum of the rounded values in the rows above. There is therefore a possibility that a columns values may not exactly sum to the value in the total cell, the total cell is however accurate, and this is only a minor rounding discrepancy with no impact on accuracy or validity. **Note that in the cost calculation exported vehicle earnings are subtracted from the total cost (reducing the total cost) so the total scenario cost is not a sum of the bottom row totals. MAC Maintenance Derogation Impact Assessment RPA | 26 Annex 3 Cost variable data tables A3.1 Scenario 1 specific data Annex Table 7: Scenario 1 ICE retirement value loss values Years of service left at Scrappage cost per car Year of ICE purchase retirement () 2027 0.5 495 2028 1.5 1,504 2029 2.5 2,553 2030 3.5 3,645 2031 4.5 4,780 2032 5.5 5,960 Total Source: RPA analysis Annex Table 8: Scenario 1 premature EV cost incurred Years of cost brought Cost per vehicle brought Year of ICE purchase forward forward () 2027 0.5 964 2028 1.5 2,854 2029 2.5 4,671 2030 3.5 6,418 2031 4.5 8,098 2032 5.5 9,714 Total Source: RPA analysis Annex Table 9: Scenario 1 MAC retirement loss values Years of service left at Scrappage cost per MAC Year of ICE purchase retirement system () 2027 0.5 21 2028 1.5 63 2029 2.5 107 2030 3.5 153 2031 4.5 200 2032 5.5 250 Total Source: RPA analysis A3.2 Scenario 2 specific data Annex Table 10: Scenario 2 ICE retirement value loss values Years of service left at Scrappage cost per car Year of ICE purchase retirement () 2027 (second half) 12 14,865 2028 12 14,865 2029 12 14,865 2030 12 14,865 MAC Maintenance Derogation Impact Assessment RPA | 27 Total cost () 846,821,147 4,527,281,055 6,767,439,524 8,498,537,242 9,790,966,353 5,383,278,815 35,814,324,134 Total cost () 1,649,523,331 8,591,085,042 12,380,453,441 14,966,115,826 16,589,257,459 8,773,773,566 62,950,208,664 Total cost () 15,110,076 80,781,594 120,753,394 151,641,875 174,703,064 96,055,411 639,045,414 Total cost () 25,448,339,626 44,753,976,584 39,401,049,973 34,662,393,629 Annex Table 10: Scenario 2 ICE retirement value loss values Years of service left at Scrappage cost per car Year of ICE purchase retirement () 2031 12 14,865 2032 12 14,865 Total Source: RPA analysis Annex Table 11: Scenario 2 premature EV cost incurred Years of service left at Scrappage cost per car Year of ICE purchase retirement () 2027 (second half) 12 18,809 2028 12 18,809 2029 12 18,809 2030 12 18,809 2031 12 18,809 2032 12 18,809 Total Source: RPA analysis Annex Table 12: Scenario 2 MAC retirement loss values Years of service left at Scrappage cost per MAC Year of ICE purchase retirement system () 2027 (second half) 12 789 2028 12 789 2029 12 789 2030 12 789 2031 12 789 2032 12 789 Total Source: RPA analysis Total cost () 30,450,254,656 13,426,192,975 188,142,207,444 Total cost () 32,200,268,118 56,628,057,725 49,854,897,880 43,858,985,885 38,529,286,335 16,988,417,318 238,059,913,261 Total cost () 574,558,760 1,010,430,924 889,575,460 782,588,657 687,489,275 303,129,277 4,247,772,353 A3.3 Shared data between scenarios Annex Table 13: CO2 emissions and value of emissions from premature EV production Year CO2 volume (tonnes) CO2 value () 2027 29,959,342 2,516,584,735 2028 52,687,119 4,425,717,982 2029 46,385,326 3,896,367,400 2030 40,806,690 3,427,761,966 2031 35,847,902 3,011,223,803 2032 15,806,136 1,327,715,395 Total 221,492,515 18,605,371,280 Source: RPA analysis Annex Table 14: HFO equipment retirement costs Year of equipment Years of service left at purchase retirement 2030 0.5 2031 1.5 Scrappage cost per machine () 120 366 Total cost () 4,570,655 13,894,791 MAC Maintenance Derogation Impact Assessment RPA | 28 Annex Table 14: HFO equipment retirement costs Year of equipment Years of service left at purchase retirement 2032 2.5 2033 3.5 2034 4.5 2035 5.5 2036 6.5 2037 7.5 2038 8.5 2039 9.5 Source: RPA analysis Scrappage cost per machine () 621 886 1,162 1,449 1,748 2,058 2,381 2,717 Total Total cost () 23,591,892 33,676,877 44,165,262 55,073,182 66,417,419 78,215,425 90,485,352 103,246,076 513,336,931 Annex Table 15: Number and value of prematurely scrapped ICE vehicles which are exported Year Cars available for export Value of exports () 2027* 2028 3,010,693 18,333,505,462 2029 2,650,590 16,140,674,416 2030 2,331,811 14,199,479,720 2031 2,048,452 12,473,973,324 2032 903,208 5,500,051,639 Total 10,944,753 66,647,684,561 Source: RPA analysis *Some vehicles do see early retirement in 2027 but for the purpose of this assessment any vehicle with less than 1 year of service left at retirement is assumed not to be exported. MAC Maintenance Derogation Impact Assessment RPA | 29 0 RPA Risk &Policy Analysts Risk & Policy Analysts Limited Suite C, 2nd floor, The Atrium, St Georges Street Norwich, Norfolk, NR3 1AB, United Kingdom Tel: E-mail: M@rpaltd.co.uk Website: www.rpaltd.co.uk If printed by RPA, this report is published on 100% recycled paper