Document VKL9OXR25wDkwaqkEngK8GbXq

ECHA EUROPEAN CHEMICALS AGENCY ANNEX XV RESTRICTION REPORT PROPOSAL FOR A RESTRICTION (DRAFT) SUBSTANCE NAME(S): Polyfluoroalkyl and perfluoroalkyl substances (PFAS) in firefighting foams CONTACT DETAILS OF THE DOSSIER SUBMITTER: VERSION NUMBER: DATE: Telakkakatu 6, P.O. Box 400, FI-00121 Helsinki, Finland I Tel. I Fax +358 9 68618210 I echa.europa.eu ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS TABLE OF CONTENTS Report...........................................................................Error! Bookmark not defined. 1. The problem identified...........................................................................................18 1.1. Hazard, exposure/emissions and risk....................................................................18 1.1.1. Identity of the substance(s), and physical and chemical properties ..................18 1.1.1.1. Overview .........................................................................................18 1.1.1.2. Long chain PFAS...............................................................................21 1.1.1.3. Short chain PFAS ..............................................................................21 1.1.1.4. Derivates of perfluoroalkyl sulfonic PFAS (also PASF-based substances) ..22 1.1.1.5. Fluorotelomers .................................................................................23 1.1.1.6. Other PFAS substances......................................................................24 1.1.1.7. Chemical definitions of the identified substances which could be used for a possible future restriction ................................... Error! Bookmark not defined. 1.1.2. Justification for grouping ............................................................................27 1.1.3. Classification and labelling..........................................................................27 1.1.4. Hazard assessment....................................................................................27 1.1.5. Exposure assessment ................................................................................28 1.1.6. Risk characterisation..................................................................................31 1.2. Justification for an EU wide restriction measure .....................................................31 1.3. Baseline............................................................................................................32 1.3.1. Overview..................................................................................................32 1.3.2. Definition of the baseline scenario for the assessment of economic impacts ......32 1.3.3. Overview of current regulatory measures .....................................................33 1.3.3.1. Stockholm Convention.......................................................................33 1.3.3.2. EU Regulation ..................................................................................34 1.3.3.3. Other international controls ...............................................................35 1.3.4. Industry Measures.....................................................................................37 1.3.4.1. Substitution and phase-out ................................................................37 1.3.4.2. Containment and control ...................................................................38 2. Impact assessment ...............................................................................................39 2.1. Introduction ......................................................................................................39 2.2. Risk Management Options...................................................................................39 i ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 2.3. Restriction scenario(s) ........................................................................................40 2.4. Economic impacts ..............................................................................................43 2.5. Human health and environmental impacts.............................................................46 2.6. Other impacts, practicability and monitorability .....................................................47 2.7. Proportionality (including comparison of options) ...................................................48 2.7.1. Comparison of different users .....................................................................48 2.7.2. Transition periods......................................................................................50 2.7.3. Concentration thresholds............................................................................51 2.7.4. Other risk management targeted at reducing release.....................................53 3. Assumptions, uncertainties and sensitivities .............................................................55 4. Conclusion ...........................................................................................................55 Annex A: Manufacture and uses .................................................................................59 A.1. Manufacture, import and export...........................................................................59 A.2. Uses.................................................................................................................59 A.3. Uses advised against by the registrants ................................................................70 Annex B: Information on hazard and risk.....................................................................71 B.1. Identity of the substance(s) and physical and chemical properties ...........................71 B.2. Manufacture and uses (summary)........................................................................99 B.3. Classification and labelling ................................................................................ 100 B.4. Environmental fate properties............................................................................ 101 B.5. Human health hazard assessment...................................................................... 102 B.6. Human health hazard assessment of physicochemical properties ........................... 106 B.7. Environmental hazard assessment ..................................................................... 106 B.8. PBT and vPvB assessment................................................................................. 106 B.9. Exposure assessment ....................................................................................... 106 B.10. Risk characterisation ...................................................................................... 137 Annex C: Justification for action on a Union-wide basis................................................ 138 Annex D: Baseline .................................................................................................. 139 Annex E: Impact Assessment................................................................................... 140 ii ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS E.1. Risk Management Options ................................................................................. 140 E.2. Alternatives..................................................................................................... 140 E.3. Restriction scenario(s) ...................................................................................... 171 E.4. Economic impacts ............................................................................................ 171 E.5. Risk reduction capacity ..................................................................................... 245 E.6. Other impacts, practicability and monitorability ................................................... 250 E.7. Practicality and monitorability............................................................................ 250 E.8. Proportionality (comparison of options)............................................................... 250 E.9.1. Comparison of Restriction Options ................................................................... 253 E.9.2. Comparison of costs and benefits .................................................................... 253 Annex F: Assumptions, uncertainties and sensitivities ................................................. 260 Annex G: Stakeholder information ............................................................................ 262 TABLE OF TABLES Table 1.1 Table 1.2 Table 1.3 Table 1.4 Table A.1 Table B.1 Table B.2 Table B.3 Table B.4 Table B.5 Table B.6 Overview of the PFAS classification, generalised chemical structures, and minimal number of C-atoms of substances that were identified as being used in AFFF 25 Comparison of the substance identification as in the PFOA restriction and a proposal made for the PFAS-substances in AFFF. 26 Overview of ratios for emissions by different environmental compartment for all life-cycle stages combined. Error! Bookmark not defined. Overview or ratios for emissions by different life-cycle stages Error! Bookmark not defined. Tonnage of fluorosurfactants purchased for the production of fire-fighting foams by manufacturers participating in the 2018 Eurofeu survey 60 Identified hydrocarbons (identified by CAS) incl. CAS/EC identifier, the substance name, chemical group and the supplier and/or product name 74 Identified detergents (identified by CAS) incl. CAS/EC identifier, the substance name, chemical group and the supplier and/or product name 77 Siloxanes (identified by CAS) incl. CAS/EC identifier, the substance name, chemical group and the supplier and/or product name 82 PFSAs (identified by CAS) with C6 incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 84 PFCAs (identified by CAS) with C7 incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 85 PFSAs (identified by CAS) with <C6 incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 86 iii ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Table B.7 Table B.8 Table B.9 Table B.10 Table B.11 Table B.1 Table B.1 Table B.1 Table B.2 Table B.3 Table B.4 Table B.5 Table B.6 Table B.7 Table B.8 Table E.1 Table E.2 Table E.3 Table E.4 Table E.5 Table E.6 Table E.7 Table E.1 Table E.2 Table E.3 Table E.4 Table E.5 PFCAs (identified by CAS) with <C7 incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 87 Identified derivates of perfluoroalkyl sulfonic PFAS (also PASF-based substances) 89 Fluorotelomer (identified by CAS) substances incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 92 Other per- or polyfluorinated substances (identified by CAS) incl. CAS/EC identifier, the designation, the acronym and the supplier and/or product name 97 Summary of key preliminary market analysis results 100 Overview of PFAS substances mobility using log Koc 101 Overview on substances used in fluorine-free fire-fighting foams and one substance used in a fluorinated foam. Shown are the product, CAS/EC, PNECs, and the used reference. The respective lowest PNECs are highlighted in bold. 103 Industry splits and usage rates based on data from Eurofeu and Brooke et al (2004)* 118 Summary of factors applied to data 120 Final selection of substances (substances highlighted in blue selected) - see also footnotes at end of table. 123 PFAS based substances for selection 127 Overview of ratios for emissions by different environmental compartment for all life-cycle stages combined. 128 Overview or ratios for emissions by different life-cycle stages 128 log Koc values for a set of solvents, POPs and PFAS based substances as indicative guide to partitioning against Koc values. 130 log Koc values for non-fluorinated substances included within this assessment 130 Shortlist of fluorine-free alternative products for assessment 142 Assessment of Respondol ATF 3-6% 144 Assessment of Re-Healing Foam RF3x6 ATC 146 Assessment of Re-Healing Foam RF1 1% 148 Assessment of Moussol FF 3x6 (F-15) 149 Assessment of FOAMOUSSE 3% F-15 151 Assessment of Ecopol Premium 152 Estimated costs for the replacement of fire extinguishers in the whole of the EU 173 Effectiveness of alternatives - summary 179 Use patterns of alternatives - summary 182 Annual foam costs - input assumptions 184 Scenarios, gross and net foam costs -annual cycle replacement costs for total EU market (M denotes millions) 185 iv ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Table E.6 Table E.7 Table E.8 Table E.9 Table E.10 Table E.11 Table E.12 Table E.13 Table E.14 Table E.15 Table E.16 Table E.17 Table E.1 Quantitative data - economic costs 186 Quantitative data - economic costs 186 Overview of key impacts of alternatives 190 Typical cost per site of remediation of PFAS contamination resulting from the use of fire-fighting foams 198 Availability of alternatives - summary 203 "Top down" assessment - annual demand and supply of PFAS and Fluorine free FFF 204 Processing time based on existing incinerator capacity processing 25l per hour. 218 Processing time based on existing incinerator capacity processing one tonne per hour. 218 Estimated costs of disposal 219 Estimate of total PFAS emissions from fire-fighting foams 246 Estimate of total quantifiable cost of a potential restriction on PFAS in fire- fighting foams 247 Estimate of cost-effectiveness of the reduction of PFAS emissions from fire- fighting foams 249 Summary of socio-economic considerations for the main expected impacts of potential regulatory management options 254 TABLE OF FIGURES Figure 1.1 Hierarchical clustering of the identified short-, long-chain and substituted PFAs substances 19 Figure 1.2 Hierarchical clustering of identified fluorotelomers 20 Figure 1.3 Split of PFAS-based fire-fighting foams by sector 29 Figure 2.1 Map summarising potential effects of a restriction on the placing on the market of PFAS-based fire-fighting foams 41 Figure 2.2 Map summarising potential effects of a restriction on the use of PFAS-based fire-fighting foams 42 Figure A.1 Split of PFAS-based fire-fighting foams by sector 62 Figure A.2 Yearly use of fluorine-free firefighting foams by sector. 68 Figure B.1 Chemical structure of sodium octyl sulphate 76 Figure B.2 Sugar-based siloxane as described by Hetzer et al. 82 Figure B.3 Hierarchical clustering of the identified short-, long-chain and substituted PFAs substances 91 Figure B.4 Hierarchical clustering of identified fluorotelomers 96 Figure B.5 Chemical structure of Dodecafluoro-2-methylpentan-3-one, a fluorinated ketone 97 v ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Overview - Problem Analysis Investigation Report Grounds for the restriction proposal and possible issues identified This Problem Analysis Investigation Report (PAIR) gives the status of the investigation to propose an Annex XV restriction on the placing on the market and use of PFAS in firefighting foams. The option of restricting in addition the manufacture of PFAS foam concentrates is also discussed. The report is intended to be a basis for discussion with the Commission on how to take forward this restriction proposal. In July 2020, the Commission requested ECHA to prepare a restriction proposal in accordance with Article 69(1) of REACH of PFAS in firefighting foam and to cooperate with the authorities working on the broad restriction of PFAS1 (hereafter referred as "universal PFAS restriction"). ECHA updated the Registry of Intentions and announced a planned submission date of 1st October 20212. The present Annex XV report is largely based on study jointly contracted by the Commission and ECHA in 2019 on possible regulatory management options to address the risks of the risks of PFAS in firefighting foams in the EU3, hereafter referred to as Wood (2020). This study included extensive literature searches, stakeholder surveys and a workshop. However, comments received from stakeholders during the opinion-making process of the PFHxA restriction proposal4 and a call for evidence held to support the development of a universal PFAS restriction proposal by NL, DE, SE, DK and NO5 (which at the time included the use of these substances in firefighting foam) indicated that certain key elements of Wood (2020) might need to be revised, in particular the annual usage rate of foams and the total stock in the EU. On this basis, additional consultancy work was contracted by ECHA early 2021 to clarify a series of elements with stakeholders and based on literature review. The outcome of this work is expected in June 2021 and will be incorporated in the present report. Since these key input parameters affect several calculations and sections of the Annex XV report, significant review and update time is envisaged to take place after the new data is received. Despite several stakeholder consultations, it is likely that some of the information sought for the analysis will not be available or will be of low quality (e g lack of supporting evidence, poor justification for representativeness, lack of specificity, large uncertainties, conflicting information) In the absence of further information from stakeholders, best estimates and robust assumptions will be generated using the information available, highlighting the limitations and uncertainties Assumptions will be revised during opinion making if additional credible information become available The necessary information to revise an assumption (i e that would be needed to be submitted during the consultation on the Annex XV report) could be clearly stated in the Annex XV report 1 https://echa.europa.eu/documents/10162/13641/request echa pfas fff en.pdf/aa089887-bc27e642-747e-b935809075cc 2 https://echa.europa.eu/registry-of-restriction-intentions/-/dislist/details/0b0236e1856e8ce6 3 https://echa.europa.eu/documents/10162/28801697/pfas flourinefree alternatives fire fighting en.pdf/d5b24e2a-d027-0168-cdd8-f723c675fa98 4 Restriction on undecafluorohexanoic acid (PFHxA), its salts and related substances available at: https://echa.europa.eu/restrictions-under-consideration/-/substance-rev/25419/term 5 https://www.reach-clp-biozid-helpdesk.de/SharedDocs/Meldungen/DE/REACH/2020-05-08RMOA-PFAS.html ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Hazard and concern PFAS are persistent substances and most PFAS are also easily transported in the environment covering long distances away from the source of their release. These substances have been frequently observed to contaminate groundwater, surface water and soil. Remediating polluted sites is technically difficult and costly. If releases continue, they will continue to accumulate in the environment, drinking water and food. Wood's study did not address in detail the hazard and risk of PFAS as this was initially planned to be developed by ECHA. In addition, considering the Commission's request to cooperate with the authorities working on the universal PFAS restriction, ECHA identified that the work performed by these authorities on the scope of substances identity, the hazard and risk could be used in ECHA's restriction proposal on PFAS in firefighting foams. This integration would benefit all parties by saving resources and experimenting the approach (sometimes termed `ice-breaker') with a restriction dossier that would undergo the RAC-SEAC opinion-making ahead of the universal PFAS restriction. The five authorities would then be able to use the learnings from the opinion-making of ECHA's restriction proposal to adapt their own proposal prior to, or in the initial period after, its submission. The main challenge with the proposed approach is the timing of ECHA's restriction proposal preparation for which final input from the authorities on the SID scope, hazard and risk would be needed by summer 2021 The authorities have shown commitment to the suggested approach and further discussions in June 2021 will take place to confirm it and the associated timeline ECHA is providing precise information on timelines and deliverables (and formats) within a `concept paper', that is currently being drafted Should it not be possible to use the hazard approach of the universal restriction, this is likely to require additional work from ECHA (possibly using a consultant) to address this part of the Annex XV report A key decision point over the next few weeks will be on whether to continue cooperation with the universal restriction or whether to invoke the contingency plan Emissions modelling In terms of emission and exposure assessment Wood (2020) chose two identified PFAS substances to calculate and represent the overall emissions of PFAS to the environment (water, air, soil) based on the quantities of firefighting foams used per sector or type of use and emission parameters across the foam life cycle based on standard parameters or expert judgement Due to the absence of specific annual usage rates for all sectors and types of use (especially training), the emission parameters are based on assumptions from the and might not represent the reality or could be challenged during the opinion-making process The shares and amounts of PFAS foams (a) still used in training and (b) never used and sent to incineration when reaching expiry date are still unclear These elements have an important impact on the emissions and the disposal costs under the restriction option where PFAS foams have to be disposed of at the end of the transitional period The latest attempt to contact stakeholders (survey targeted on the users) has been in initiated on 6 May 2021 to clarify these elements (deadline for response: 4 June 2021) In absence of more specific information, assumptions will be made based on the more generic information already available Also, the approach taken for the emission characterisation is simple and does not model complex fate aspects (e g flow across environmental compartments, degradation 7 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS products, direct human exposure) This simple assessment has been chosen based on RAC and SEAC approach to consider emissions as a proxy for risks and impacts (`PBT/vPvB approach') and due to the complexity of modelling the fate of a very large number of PFAS substances This approach could also be challenged during the opinion-making process unless RAC agrees that PFAS should be considered as non-threshold and that releases should be minimised as per PBT/vPvB substances (see links to hazard/risk above) Independently of this issue, the emission estimation will model the variation of emissions per sector of use according to different transitional periods Cost calculations The cost calculations consider both costs and savings to industry These comprise price differences between PFAS and fluorine-free foams, higher quantities of fluorine-free foams needed to achieve the same level of fire-protection, savings in clean-up and incineration costs The absence of clear information on some of the main input data affecting cost calculations, especially the amount of foam stocks in the EU and the cost of equipment cleaning (to achieve different concentration limits) and change in equipment to use alternatives, leads to large ranges of costs associated with the restriction options However, we are currently trying to collect more specific information which would allow narrowing down these cost ranges The possibility to assess technical RMMs to contain releases for sectors where containment may be possible cannot be currently assessed from the cost perspective due to lack of data, however, we are currently collecting information on that as well 8 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Seveso sites management would be eligible measures are implemented (containment and treatment), these are never 100% effective, especially in case of large fire incidents. Main advantage: allows taking into account site- specific constraints for implementing the substitution. 5 Restriction for all RMO proposed uses unless based on the adequate risk discussions on the management PFHxA restriction. measures are in Main advantage: place to minimise allows substitution the emissions to if technically and the environment economically feasible and continued use of PFAS foams where this is not feasible, to the condition that adequate containment and reduction of emissions is ensured. Incentive for industry to provide additional information on conditions of use. However, in practice, this is likely to be only applicable in certain sites being able to implement such strict RMMs. Even if risk management measures are implemented (containment and treatment), these are never 100% effective, especially in case of large fire incidents. 13 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 1. The problem identified 1.1. Hazard, exposure/emissions and risk 1.1.1. Identity of the substance(s), and physical and chemical properties 1.1.1.1. Overview Polyfluoroalkyl and perfluoroalkyl substances (PFAS) is a broad term used to cover approximately 4 700 substances. Long chain (C8) PFAS were used as surfactants specifically because of their potent water and oil repellence at low concentrations (Buck, 201110). However, short chain PFAS are nowadays used since the phase out of the long chain substances. According to Eurofeu, PFAS used in firefighting foam technology presently exclusively are of the C6-type.11 The main function of PFAS in fire-fighting foam is to act as a surfactant, that is to form a film over the surface of a burning liquid in order to prevent flammable gases from being released from it as well as reigniting. They are therefore used in fires involving flammable liquid (Class B fires12) across a range of sectors. The quantities of foam used by different sectors are discussed in the sub-section below. PFAS-based fire-fighting foams are used for fires in many different applications involving flammable liquids,and are used in equipment such as small fire extinguishers up to large tank fires. They can be applied with both mobile and stationary equipment and are also used in training and testing of equipment. A large number of highly diverse PFAS substances have been identified as being used in fire-fighting foams, as follows: Unsubstituted13 long chain PFAS (see Section 1.1.1.2 and Annex B 1.1.1 and B 1.1.2) Unsubstituted short chain PFAS (see Section 1.1.1.3 and Annex B 1.1.3 and B 1.1.4) Substituted short and long chain PFAS (see Section 1.1.1.4 and Annex B 1.1.5) 10 Buck, R. C., Franklin, J. , Berger, U. , Conder, J. M., Cousins, I. T., de Voogt, P. , Jensen, A. A., Kannan, K. , Mabury, S. A. and van Leeuwen, S. P. (2011), Perfluoroalkyl and polyfluoroalkyl substances in the environment: Terminology, classification, and origins. Integr Environ Assess Manag, 7: 513-541. doi:10.1002/ieam.258 11 Eurofeu (2020), "Comments on the PFHxA-Restriction Proposal", comment #2983 to the PFHxA restriction proposal, available in the document "part 2" available at: https://echa.europa.eu/restrictions-under-consideration/-/substance-rev/25419/term 12 The European Standard Classification of Fires distinguishes between the following fires: Class A - fires involving combustible solid materials (e.g. wood, paper or textiles); Class B - fires involving flammable liquids (e.g. petrol, diesel or oils); Class C - fires involving gases; Class D - fires involving metals; Class K - fires involving live electrical apparatus; Class F - fires involving cooking oils. 13 PFAS substances with perfluorinated alkaline carbon chain that is connected to a sulfonic- or carboxylic acid head group, not further equipped with additional chemical groups. 18 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Fluorotelomers (see Section 1.1.1.5 and Annex B 1.1.6) Others (See Section 1.1.1.6 and Annex B 1.1.7) Summary overviews of the long- and short-chain PFASs identified and of the fluorotelomer substances identified are provided in the figures below. The figures below provide a hierarchical clustering of the substances that have been identified as being used in fire-fighting foams, including short-, long-chain and substituted PFAS (Figure 1 1) and also fluorotelomers (Figure 1 2). Figure 1 1 Hierarchical clustering of the identified short-, long-chain and substituted PFAs substances 19 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 1 2 Hierarchical clustering of identified fluorotelomers 20 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 1.1.1.2. Long chain PFAS The group of long chain PFAS consists of both perfluorosulfonic acids (PFSAs) with C6 (chemically defined as CnF2n+1-SO3H)14 and perfluorocarboxylic acids (PFCAs) with C815 (chemically defined as CnF2n+1-COOH). The group of long chain PFAS (defined by OECD as perfluorosulfonic acids (PFSAs) with C6 and perfluorocarboxylic acids (PFCAs) with C8) encompasses the following substances: PFSAs with C6 (see a full overview of these substances in Annex B 1.1.1) Perfluorohexane sulfonic acid (PFHxS) (C6) Perfluoroheptane sulfonic acid (PFHpS) (C7) Perfluorooctane sulfonic acid (PFOS) (C8) Perfluorononane sulfonic acid (PFNS) (C9) Perfluorodecane sulfonic acid (PFDS) (C10) Perfluoroundecan sulfonic acid (PFUnDS) (C11) PFCAs with C8 (see a full overview of these substances in Annex B 1.1.2) Perfluorooctanoic acid (PFOA) (C8) Perfluorononanoic acid (PFNA) (C9) Perfluorodecanoic acid (PFDA) (C10) Perfluoroundecanoic acid (PFUnDA) (C11) Perfluorododecanoic acid (PFDoDA) (C12) Perfluorotridecanoic acid (PFTrDA) (C13) Perfluorotetradecanoic acid (PFTeDA) (C14) Perfluorostearic acid (PFOcDA) (C18) 1.1.1.3. Short chain PFAS Short chain PFAS identified include sulfonic acids and carboxylic acids: PFSAs with C2-C5 (see Annex B 1.1.3) Perfluoroethane sulfonic acid (PFEtS) (C2) Perfluoropropane sulfonic acid (PFPrS) (C3) Perfluorobutanesulfonic acid (PFBS) (C4) 14 The PFSAs and PFCAs identified as used in FFFs are covered by the general classificat on of per- and polyfluoroalkyl substances (PFASs) by the OECD. 15 C8 covers in this case perflouralkylcarbons as well as the single carbon belonging to the ac d group. 21 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Perfluoropentane sulfonic acid (PFPeS) (C5) PFCAs with C4-C6 (see Annex B 1.1.4) Perfluoro-n-butanoic acid (PFBA) (C4) Perfluoropentanoic acid (PFPeA) (C5) Perfluorohexanoic acid (PFHxA) (C6) Perfluoroheptanoic acid (PFHpA) (C7) 1.1.1.4. Derivates of perfluoroalkyl sulfonic PFAS (also PASF-based substances) All the named substances above are characterised by a perfluorinated alkaline carbon chain that is connected to a sulfonic- or carboxylic acid head group. In other PFAS substances, this head group is also equipped with additional chemical groups. These other substances form the group derivates of perfluoroalkyl sulfonic PFAS, also called perfluroalkane sulfonyl fluoride substances (PASF), as their synthesis is based on perfluroalkane sulfonyl fluoride. This can be for example an amide (sometimes methylated or ethylated). The chemical formulae of this group can be summarised as: Perfluoroalkane sulfonyl fluoride (PASF) = CnF2n+1SO2F PASF-based derivates = CnF2n+1SO2-R, where R = NH, NHCH2CH2OH, etc. However, in most cases, these substances were not found when the actual foam was tested but rather when environmental samples were tested. In addition, some of the substances are also known to be environmental transformation products. Other substances are raw materials for surfactant and surface protection products (EtFOSE and N-MeFOSe) (Buck et al. 2011). In this sub group, the following substances were found: Perfluoroalkyl sulfonamido amines (PFOSaAm); C7-FASA - Perfluoroheptane sulfonamidoethanol (PFHpSA); C8-PFSiA (PFOSI); N-Ethyl perfluorooctane sulfonamidoacetic acid (EtFOSAA); N-Methyl perfluorooctane sulfonamidoacetic acid (EtFOSE); N-Methylperfluorobutanesulfonamide (FBSA); Perfluorooctane sulfonamide (FOSA); Perfluorooctane sulfonamido acetic acid (FOSAA); Perfluorooctane sulfonamidoethanol (FOSE); N-methyl perfluorooctanesulfonamidoacetic acid (N-MeFOSA); N-Methyl perfluorooctane sulfonamidoethanol (N-MeFOSE); Perfluoroalkyl sulfonamido amines (PFBSaAm); N-[3-(Dimethyloxidoamino)propyl] -3,3,4,4,5,5,6,6,7,7,8,8,8-Tridecafluor-1- octanesulfonamid; and Carboxymethyl)dimethyl [3- (gamma-omega-perfluor-1-C6-14- Alkansulfonamid)propyl)ammonium. 22 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS A full overview of the substituted short and long chain PFAS identified is provided in Annex B 1.1.5. 1.1.1.5. Fluorotelomers Fluorotelomers are defined by having an additional non-fluorinated spacer between the perfluorinated alkyl chain and the charged head group (denotated as number of perfluorinated carbons: number of non-fluorinated carbons). The most known homologues of this subgroup are those that have a two carbon atom spacer (defined as CnF2n+1-C2H4R)16. Fluorotelomers cover a wide range of positively/negatively charged head groups or combinations of those. Most of the fully identified substances, exhibit the xx:2 structure, where two non-fluorinated carbon atoms are inserted between the perfluorinated carbon chain and the head group. However, in the case of fluorotelomer betaines also xx:1:2 and xx:3 are found. In the latter case, three non-fluorinated carbon atoms are inserted between the perfluorinated carbon chain and the head group. In the case of the xx:1:2 substances, an additional fluorinated carbon is inserted between the perfluorinated alkyl chain and the non-fluorinated spacer. Based on the manufacturing dates that are cited in the respective publications, it can be assumed that the use of fluorotelomers in fire-fighting foams began later than the use of traditional PFAS substances without a non-fluorinated spacer. The following head groups have been identified: Alkylbetaine (AB); Betaine (B); Carboxylic acid (CA); Hydroxy (OH); Thioamido sulfonates (TAoS); Unsaturated carboxylic acid (UCA); Sulfonamido betaines (SaB); Sulfonamide amine (SaAm); and Thio hydroxy ammonium (THN+). A full overview of the fluorotelomer substances identified is provided in Annex B 1.1.6. 16 This corresponds with the general classification of per- and polyfluoroalkyl substances (PFASs) by the OECD, however flurotelomers used in FFFs have also been identified with a spacer of three non fluorinated carbon atoms (for example 7:3 FTB), as well as fluorotelomers with a nonfluorinated and an additional single-fluorinated carbon. 23 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 1.1.1.6. Other PFAS substances In some cases, perfluorinated substances that do not belong to any of the named groups (long-/short-chain PFAS, fluorotelomers, and derivates of perfluoroalkyl sulfonic PFAS) have been identified (`others'). The identified substances grouped under the term "others" show diverse chemical structures. A full overview of the chemical substances attributed to `other' is provided in Annex B 1.1.7. The only feature that is common to all of them is a perfluorinated substructure. The substances include: Ammonium propionate 2,2,3 trifluor-3-(1,1,2,2,3,3-hexafluoro-3-trifluormethoxypropoxy), Dodecafluoro-2-methylpentan-3-one Poly(1,1,2,2-tetrafluoro-1,2-ethanediyl), -fluoro--2-[3-((carboxylatomethyl) dimethylammonio)propylaminosulfonyl]ethyl- Thiols, C8-20, gamma-omega-perfluoro, telomers with acrylamide Sodium p-perfluorous nonenoxybenzene sulfonate Bis(trifluorovinyl)ether However, in analogy to the perfluorinated ethers like ADONA (CnF2n+1-O- CmF2m+1), the substance Dodecafluoro-2-methylpentan-3-one (a ketone) could be defined by CnF2n+1-COCmF2m+1. 1.1.1.7. Substance identity restriction scope Any regulatory action on chemical substances/substance groups relies on a precise chemical identification. In the following the identified PFAS-substances have been checked to confirm whether they are covered by the general classification of per- and polyfluoroalkyl substances (PFASs) by the OECD, which itself is based on the commonly agreed terminology for nomenclature of PFASs (Buck et al. 2011). In the case of the PFCAs, chemically defined as CnF2n+1-COOH, the substances identified in this task, AFFF-related PFAS-substances, would be covered. This is also true for the sulfonic homologues (PFSA, defined as CnF2n+1-SO3H). Fluorotelomers-based substances are chemically defined by having a non-perfluorinated spacer between the perfluorintated carbon chain and a polar head group. The most known homologues of this subgroup are those that have a two carbon atom spacer (defined as CnF2n+1-C2H4-R). This definition is also given in the OECD report ("Working towards a global emission inventory of PFASs: focus on PFCAS - status quo and the way forward"). In this task, multiple substances belonging to this group were identified, varying in the perfluorinated chain length. However, homologues with a spacer of three non fluorinated carbon atoms (for example 7:3 FTB) were also identifed, thus the definition would need to be enlarged to CnF2n+1CmHm+1-R, so that substance with a variable chain length could be included. In addition fluorotelomers with a non-fluorinated and an additional single-fluorinated carbon were identified (for example 7:1:2 FTB). In order to include such substances, the chemical definition for these homologues would need to be (CnF2n+1-CHF-CmHm+1-R). 24 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS oil repellence21. The same properties mean that many PFAS substances are also highly mobile (within the natural environment) and highly persistent (see Appendix 3 of the underlying study22). This can create issues where PFAS substances emitted to the environment reach and contaminate important resources such as groundwater23. There is evidence to suggest that exposure to PFAS can lead to adverse health effects in humans (by eating or drinking food or water contaminated by PFAS). In particular the US EPA24 highlight studies that indicate the longer chain (C8 PFAS) species PFOS and PFOA can cause reproductive and developmental, liver and kidney, and immunological effects on laboratory animals. Furthermore, both chemicals have caused tumours in animal studies. Their use is already restricted in the EU and elsewhere. Some short-chain PFAS (PFHxS, PFBS, HFPODA) have also been listed as SVHCs, based on there being an equivalent level of concern to the named groups of chemicals under the authorisation provisions under REACH (carcinogens, mutagens and reprotoxicants (CMRs) and persistent, bioaccumulative and toxic/very persistent and very bioaccumulative (PBTs/vPvBs) chemicals). The Nordic Council of Ministers25 commented that the annual health-impacts within an EEA exposure study (from all uses of PFAS, not only fire-fighting foams) was estimated at 5284 billion. This gives an indication of the scale of the issue and magnitude of the potential impacts from the environmental build-up of PFAS. The same study describes remediation costs associated with contamination from PFAS at European sites ranging from several hundred thousand up to 40 million with one high-cost example for the Dusseldorf Airport, Germany estimating a total remediation cost of up to 100 million. Based on the physical properties of PFAS (particularly mobility and persistence) along with identified health effects for some PFAS, PFAS represent a challenging environmental and human health hazard issue. 1.1.5. Exposure assessment Based on an extrapolation of data provided by Eurofeu (see Annex A for more details) it is estimated that around 20,000 tonnes of PFAS-based fire-fighting foams are sold in the EU per year. Of these, about 12,000 tonnes are estimated to be employed in fixed 21 Buck et al, 2011, `Perfluoroalkyl and polyfluoroalkyl substances in the environment: Terminology, classification and origins', Integrated environmental assessment and management vol 7 issue 4. 22 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commiss on DG Environment and European Chemicals Agency (ECHA) under specif c contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 23 E.g. EFSA, 2012. Perfluoroalkylated substances in food: occurrence and dietary exposure. EFSA J. 10, 2743. Available at: https://www.efsa.europa.eu/efsajournal/pub/2743; Hu et al. 2016 Detect on of Poly- and Perfluoroalkyl Substances (PFASs) in U.S. Drinking Water Linked to Industrial Sites, Military Fire Training Areas, and Wastewater Treatment Plants, Environ. Sci. Technol. Lett. 2016, 3, 10, 344-350; Hurley et al, 2016 Preliminary Associat ons between the Detection of Perfluoroalkyl Ac ds (PFAAs) in Drinking Water and Serum Concentrations in a Sample of California Women, Environ. Sci. Technol. Lett. 2016, 3, 7, 264-269; Ingelido et al, 2018, Environment Internat onal, Volume 110, January 2018, Pages 149-159 24 US EPA, 2019, `Basic information on PFAS', https://www.epa.gov/pfas/basic-information-pfas 25 Nord c Council of Ministers, 2019, `The Cost of Inact on - A soc oeconomic analysis of environmental and health impacts linked to exposure to PFAS', http://norden.divaportal.org/smash/get/diva2:1295959/FULLTEXT01.pdf 28 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS based foams in training has been reported). Stakeholder input did not allow to conclude on their relative effectiveness. In conclusion, it has been demonstrated that the use of PFAS in fire-fighting foams is associated with a significant environmental concern that does not seem to be adequately addressed by the current measures in place (current measures are discussed in more detail in Section 1.3). Even if additional measures were introduced at Member State level (and the consultation has not raised anything suggesting that they will be), there is potential for discrepancies in the definitions and scope of any national restrictions (e.g. definition of substances covered, uses covered, concentration thresholds, transition periods). This has implications not only for the degree to which the environment is protected, but also in terms of ensuring the functioning of the internal market. Firefighting foams being traded over the borders, different restrictions in different Member States could make it very challenging to market fire-fighting foam products saleable in all Member States. It would therefore not be meaningful or possible to restrict them nationally due to internal market considerations. Moreover, due to their high mobility and persistence as well as their proven ubiquity (at least of some PFAS), it appears very likely that PFAS emissions lead to crossborder pollution. Therefore, potential further regulatory management on EU-level is likely required. 1.3. Baseline 1.3.1. Overview The baseline presented here comprises an overview of the current use of PFAS-based firefighting foams based on the market analysis (used in particular as baseline economic activity for the assessment of economic impacts) and an overview of the current regulatory and voluntary industry measures to control the risk of this use. Resulting baseline exposure have already been presented in Section 1.1.5 and are not repeated here. 1.3.2. Definition of the baseline scenario for the assessment of economic impacts The baseline scenario describes the situation in the absence of any further regulatory management options (RMOs). It reflects the current market situation, but also any anticipated changes in the absence of the proposed RMOs. It was used to compare restriction scenarios (defined in the next sub-section), to ensure that the SEA evaluates the impacts of the RMOs being assessed. More details are provided in the market analysis (see Annex A), but the key points are below. It is estimated that currently some 14,000-20,000 tonnes (likely closer to the upper end of the range) of PFAS-based fire-fighting foams are sold per year in the EU and used in various sectors including chemicals/petrochemicals, municipal fire-fighting, marine, airports, military, railways and fire extinguishers. Their use is particularly important and widespread where there is a risk of Class B fires, i.e. where flammable liquids are present. They are used for fire-fighting, but in some cases also for training and testing of equipment. 32 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Some 9,000 tonnes per year of fluorine-free foams are already used in most of the same applications, although the split by sector varies from that of PFAS-based foams. Several stakeholders, including manufacturers of fire-fighting foams, have indicated that the use of fluorine-free foams has been increasing, particularly in applications where PFAS-based foams can be very easily replaced (e.g. training). This trend is expected to continue in the future to some extent (even in the absence of any restriction on PFAS-based foams). Some stakeholders also noted that containment of fire-water run-off, particularly from training, has been increasing and that this has likely reduced emissions of PFAS significantly. In addition, there are significant existing stocks of PFAS containing foams which have been already purchased. These may need to be disposed of and replaced. The total quantum of these stocks is uncertain, but are estimated as follows: o Annual sales of PFAS-based foams are estimated at between 14,000-20,000 tonnes per year. o Current annual sales of fluorine-free foams are estimated at 7,000-9,000 tonnes per year. Historically, this demand would have been served by PFAS containing foams, hence the total annual sales of PFAS-based foams could have been some 21,000-29,000 tonnes. o The shelf life of PFAS-based foams is reported to be typically between 10 and 20 years (and up to a maximum of 30 years)31. Given that foams may be used before the end of their shelf life, the actual lifetime of foams could be shorter. BiPRO 2010 suggests that the average lifespan of fire-fighting foams is 15 years, which appears consistent with the above information32. o A stock of PFAS foam of 62 500 tonnes in the EU is estimatedGiven that between 14,000 and 29,000 tonnes of PFAS based foam have historically been replaced per year, and assuming an average lifespan of foams of 15 years, indicates that the existing European stocks of PFAS based foam may be between 210,000 and 435,000 tonnes33. These volumes of stock are used in the SEA calculations. 1.3.3. Overview of current regulatory measures 1.3.3.1. Stockholm Convention The Stockholm Convention on Persistent Organic Pollutants (POPs) includes restrictions on the production and use of a number of specific PFAS, at international level, including some 31 Proposal for a restr ction: Perfluorohexane sulfon c acid (PFHxS), its salts and PFHxS-related substances https://echa.europa.eu/documents/10162/a22da803-0749-81d8-bc6d-ef551fc24e19 32 BiPRO, 2010, Study on waste related issues of newly listed POPs and cand date POPs 33 Wood (2020) estimated a stock between 210 000 and 435 000 tonnes, however, comments from stakeholders on the PFHxA restr ction proposal indicate that the figure of 62 500 tonnes would be more realistic (e.g. comment #3010 from FFFC available at in Part 2 document available at: https://echa.europa.eu/restr ct ons-under-consideration/-/substance-rev/25419/term) A lifespan of 15 years means that each year, 1/15 of the stocks are replaced. So, if between 14,000 to 29,000 tonnes are replaced per year, then the stock is 15 times that tonnage. Multiplying annual replacement tonnages w th 15 yields the above estimates. 33 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS provision for exemptions for the production and use of these compounds for use in firefighting foams. PFOS, its salts and PFOSF are listed under Annex B of the Stockholm Convention, which restricts production and use to specified acceptable purposes and specific exemptions. Upon its initial listing in 2009, an acceptable purpose was put in place for PFOS used in firefighting foams. At the POPRC meeting in 2018, the committee recommended, based on the findings of an assessment of alternatives to PFOS34 , that the acceptable purposes for the production and use of PFOS, its salts and PFOSF for fire-fighting foam be amended to a specific exemption for the use of fire-fighting foam for liquid fuel vapour suppression and liquid fuel fires (Class B fires) already in installed systems, including both mobile and fixed systems, and with the same conditions put in place for PFOA (see below). This exemption was agreed accordingly at the Ninth Meeting of the Conference of the Parties (COP) to the Stockholm Convention in 2019. At the 14th meeting of the POPRC in September 2018 the POPRC recommended listing PFOA, its salts and PFOA-related compounds in Annex A to the Convention with specific exemptions. One exemption specified was for use of firefighting foams containing PFOA already installed in systems including both mobile and fixed systems with specific conditions. Parties to the Convention can register for this exemption if they: i) ensure that FFFs that contain or may contain PFOA shall not be exported or imported except for the purpose of environmentally sound disposal; ii) do not use FFFs that contain or may contain PFOA for training or testing (unless all releases are contained) purposes; iii) by the end of 2022 if possible, but no later than 2025, restrict uses of FFFs that contain or may contain PFOA, to sites where all releases can be contained; iv) ensure all fire water, waste water, run-off, foam and other wastes are managed. This was also agreed accordingly at the 9th COP in 2019. At its fifteenth meeting, the POPRC adopted the risk management evaluation on perfluorohexane sulfonic acid (PFHxS), its salts and PFHxS-related compounds and recommended to the Conference of the Parties that it consider listing the chemicals in Annex A to the Convention without specific exemptions. The listing will not be officially adopted until the next COP meeting in May 2021, and would be officially added to the Annexes of the Convention in 2022. 1.3.3.2. EU Regulation The European Union has implemented the POPs Regulation (EC 2019/1021)35 which acts to implement the provisions of the Stockholm Convention across the EU Member States. PFOS was originally included in the restricted substances list of REACH. However, since its addition to the Stockholm Convention in 2009, it has been regulated under the POPs Regulation. PFOS, its salts and PFOSF are listed under Annex I of the POPs Regulation, specifying the following exemptions for unintentional trace contaminants (UTC)36: Substances or preparations (<10 mg/kg) 34 UNEP/POPS/POPRC.14/INF/8 : http://chm.pops.int/TheConvent on/POPsReviewCommittee/Meetings/POPRC14/Overview/tabid/7398/Default.as px 35 https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019R1021&from=EN 36 There is an exemption for the use in hard chromium plating, although that is not relevant for fire-fighting foams. 34 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS PFASs being introduced into Australia. Based on the level of toxicity and environmental persistence, the National Industrial Chemicals Notification and Assessment Scheme (NICNAS) recommends restrictions on how these substances can and cannot be used45. 1.3.4. Industry Measures 1.3.4.1. Substitution and phase-out As noted in several documents under the Stockholm Convention, for over a decade, a number of alternatives to the use of C8-based fluorosurfactants (containing PFAS) in firefighting foams have been developed and are now widely available. These include shorterchain (C6) fluoro-surfactants, as well as fluorine-free fire-fighting foams; and other developing fire-fighting foam technologies that avoid the use of fluorine. The use of C8-based AFFF has been largely phased out in favour of these alternatives. For example, it is reported that the volume of AFFF-containing PFOS used in the USA declined from around 21 million litres in 2004 to less than 9 million litres in 201146. The POPRC officially recognises that a transition to the use of short-chain per- and polyfluoroalkyl substances (PFAS) for dispersive applications such as fire-fighting foam is not a suitable option from an environmental and human health point of view and that some time may be needed for a transition to alternatives without PFAS (POPRC-14/3). In the USA, in 2006, the US EPA launched the PFOA Stewardship Program following concerns raised about the impact of PFOA and long-chain PFASs on human health and the environment, including concerns about their persistence and presence in the environment47. The programme involved eight major companies48 committing to reducing PFOA from facility emissions and product content by 95 percent no later than 2010, and to work toward eliminating PFOA from emissions and product content no later than 2015. All participating companies state in the most recent progress reports, that they met the PFOA Stewardship Program goals49. In Australia, it has been reported that the Department of Defence commenced phasing out its use of PFOS and PFOA-containing fire-fighting foams and switched to `Ansulite', which only contains trace elements of PFOS/PFOA and is only used in emergency situations or in controlled environments to test equipment. Furthermore, PFAS use is also limited by Air Services Australia, a government-owned corporation that provides air traffic control management, which has transitioned away from fluorinated firefighting foam to nonfluorinated firefighting foam including the destruction of remaining stockpiles50. 45 https://www.oecd.org/chemicalsafety/portal-perfluorinated-chemicals/countryinformat on/australia.htm 46 FFFC (2011) Estimated Inventory Of PFOS-based Aqueous Film Forming Foam (AFFF). 2011 update to the 2004 report ent tled "Estimated Quantities of Aqueous Film Forming Foam (AFFF) In the United States". Prepared for the Fire Fighting Foam Coal t on, Inc. 47 https://www.epa.gov/assessing-and-managing-chem cals-under-tsca/fact-sheet-20102015-pfoastewardship-program 48 Arkema, Asahi, BASF, Clariant, Daikin, 3M/Dyneon, DuPont, Solvay Solexis 49 https://www.epa.gov/assessing-and-managing-chem cals-under-tsca/20102015-pfoa-stewardship-program2014-annual-progress 50 https://www.oecd.org/chemicalsafety/portal-perfluorinated-chemicals/countryinformat on/australia.htm 37 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 1.3.4.2. Containment and control In Germany51, it is reported that the regulatory authorities and fire-fighting associations have compiled a leaflet on PFAS in fire-fighting, which has reportedly resulted in an increased awareness of the risks associated with certain PFASs by industry, NGOs and the public. In Norway52 it is reported that fluorine-containing fire-fighting foam has been substituted with fluorine-free alternatives in most civil airports and fluorine-containing foam is no longer in use at fire-fighting training sites with the Norwegian military forces. Furthermore, it is reported that PFAS are being gradually substituted with fluorine free-alternatives in the offshore sector, and the volumes of fluorine-containing foam used in this sector are decreasing. One respondent to the consultation questionnaire conducted for this project reported that the Swedish Petroleum and Biofuels Institute has previously (2011) provided guidance on how to plan and implement the prevention of spillage and secondary containment embankments, methods for emergency response, and for the assessment and preventing product tanks to lift off inside water filled bunds/embankments. It was estimated that ~80 % of the member companies were in compliance with this guidance. The trade association, the Fire Fighting Foam Coalition (FFFC) has published a best practice guidance document for the safe use of firefighting foams for Class B fires53, with the aim to "foster use of foam in an environmentally responsible manner so as to minimize risk from its use". The guidance covers the following aspects of Class B firefighting foam use: Foam Selection - specifying situations where the use of Class B foams is, and is not, recommended, e.g. limiting the use of Class B foams to situations that present `a significant flammable liquid hazard'. Eliminating Foam Discharge - noting that this is not always possible in emergency situations, but emphasising the possibility to achieve this in training and the testing of foam systems and equipment. Training - providing guidance on the formulation of training foams, the design, construction and operation of training facilities. Foam System Testing - including guidance on acceptance tests, conducted pursuant to installation of the system; and maintenance tests (i.e. of firefighting vehicles). Containing Foam Discharge - guidance to prevent discharge to the environment, both for manual and fixed systems. Firewater and foam concentrate disposal - with an emphasis on incineration but also covering coagulation, flocculation, electro-flocculation, reverse osmosis, and adsorption on granular activated carbon (GAC). 51 https://www.oecd.org/chemicalsafety/portal-perfluorinated-chemicals/countryinformat on/germany.htm 52 https://www.oecd.org/chemicalsafety/portal-perfluorinated-chemicals/countryinformat on/norway.htm 53 Covering aqueous film-forming foam (AFFF), alcohol resistant aqueous filmforming foam (AR-AFFF), filmforming fluoroprotein foam (FFFP), alcohol resistant film-forming fluoroprotein foam (AR-FFFP), and fluoroprotein foam (FP, FPAR). 38 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Companies handling flammable liquids in a variety and/or quantity for which the continued use of AFFF foam extinguishing agents is recommended or even existential are typically subject to the so-called SEVESO III Directive ((EU) 2012/18) concerning the control of major accident hazards involving dangerous substances According to this directive, the handling of these substances requires an official permit, which already includes fire protection measures The permit is issued on the basis of defined application forms, a defined procedure after examination and is reviewed on a recurring basis within the framework of the already existing monitoring systems by on-site inspections of the competent authorities in accordance with Article 20 of the SEVESO III Directive We propose to extend this operating license according to (EU) 2012/18 to the effect that companies can apply for an exemption for the temporary continued use of AFFF foam extinguishing agents and also receive it if it can be proven that the corresponding requirements are met These conditions could include, but are not limited to: - A detailed justification for why fluorine-free extinguishing agents cannot be switched to immediately and AFFFs should continue to be used -Evidence of appropriate containment systems and disposal routes for all fluorinecontaminated extinguishing water - Regular detailed reports on the quantities of foam agents held, procured and used, as well as proof of disposal for unused AFFF foam agents that have been replaced - A detailed transition plan to fluorine-free extinguishing agents with verifiable targets In establishing these targets, an effective combination of technical equipment and extinguishing agents must be the essential basis for the timing The requirements could be adapted to the respective state of the art/state of the general changeover, and a permit issued thereafter would be limited in time (e g , to two to three years in line with the intervals of on-site inspections under the Major Accidents Ordinance) This would - avoid attaching unrealistic and difficult-to-follow conditions to the exemptions (e g the exemption of tanks >500 square meters without taking their containment areas into account) - Companies with particularly complex fire protection systems or for which an ad hoc conversion would be particularly difficult economically, were given the opportunity to adapt the conversion to their performance capacity - In return, the authorities are given the opportunity to obtain very precise information on the material flows via the application data, to flexibly adapt the approval requirements to the general state of development, and thus to be able to phase out fluorine technology at maximum speed with minimum risk - In addition, the existing monitoring system under Article 20 of the SEVESO III Directive provides the authorities with an overview of the foam extinguishing agents held, used and procured during the recurring on-site inspections at the plants." 44 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Only very limited data for long-term success of cleaning procedures. For fire brigade vehicles and for stationary fire extinguishing system Arcadis' method indicates reaching final water rinse containing around 0.1 g/l (0.1 ppb) PFAS as measured for the sum of 28 PFAS (according to TOP 4 g/l) but no cost information was available. The PerfluorAd technique would lead to less than 1.0 g/l (1ppb) total PFAS, even often less than 0.3 g/l (0.3 ppb) with 20 000 to 25 000 per vehicle. The LfU method could reach levels as low as 10 ng/l (10 ppt) of each of the 13 standard PFAS (according to German DIN standard) with costs of 100 000 to 200 000 per vehicle. For vehicles and tanks, the methods proposed by FRV and FPA Australia could lead to levels of 70 ppt as measured by sum of PFOA and sum of PFHxS and PFOS, however, no cost information was provided. WFVD indicates fire brigade vehicles cleaning reaching levels around the g/l of each of the 13 standard PFAS (according to German DIN standard) with costs of 4 000 per vehicle. 2.7.4. Firefighting foams disposal methods The incineration of PFAS-containing firefighting foams is the most used disposal method. Literature indicates that waste incinerators at temperatures of 900 C are able to destroy PFOS at more than 99%. A destruction efficiency of more than 99.97 % for fluorotelomers, chlorofluorocarbons and PTFE in conventional waste incineration was also reported. However, this process might not lead to the complete mineralisation of the PFAS i.e. the decomposition of the PFAS to CO2, water, and hydrogen fluoride. At these temperatures short chain fluorinated compounds such as CF4, C2F6, CHF3, C2H2F2 and C3F8 can be formed and released to the air. Literature indicates that temperatures of at least 1,400 C are needed to destroy CF4 and as such completely mineralise the PFAS. Literature indicates, that 1,100 C is sufficiently hot and feasible for the destruction of PFAS, however no study has provided quantitative results on possible fluorinated gas emissions. The average cost of approximately 1/l is comparatively cheap but the process requires high amounts of energy as the water needs to be vapourised. The co-incineration of PFAS waste in cement kilns is a viable alternative to incineration in HWI, as these kilns reach temperatures of up to 1,800 C with residence times of ~20 seconds. It has been shown that the addition of calcium fluoride can increase the quality of the clinker. Additionally, calcium salts can decrease the decomposition temperature of PFAS and increase the mineralisation rate by forming calcium fluoride. Through the addition of PFAS-containing waste to the clinker production in-situ calcium fluoride can be formed, which can increase the clinker quality and destroy the PFAS. The applicability of liquid AFFF concentrate in the cement kilns is yet unknown. More detailed information on disposal methods for firefighting foam concentrates is available in section XXX of the Annex. However, stakeholders from Germany indicated, that cement kilns don't have the same filter techniques as HWIs (DUS-Valentin-Interview 2021). In addition, the cement could also be contaminated (LASTFIRE-Interview 2021). 52 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 2.7.5. Disposal of PFAS-contaminated (fire run off and cleaning) water The following conclusions can be made for available disposal options for PFAS-contaminated (fire run off and cleaning) water: Fire run off and cleaning water are highly PFAS-contaminated compared to for example groundwater techniques. Based on this not all available remediation techniques for groundwater can be used also for run off and cleaning water. Granulated activated carbon (GAC) can be also used for all PFAS-contaminated for run off and cleaning water. However, the efficiency is lower for PFAAs (carboxylic acid) in general and short chain PFAS. For other PFAS (e.g. zwitterionic) no data is available. One stakeholder brought up that when GAC is reactivated (using 800 C) PFAS could be emitted. Ion exchange (IX) is generally suited for PFAS-contaminated run off and cleaning water. However, no caste studies are available. Based on the type of PFAS various IXmatrices are available. IX is believed to be 4 times more expensive that GAC. GAC and IX are generally based on column beds to which PFAS absorb. To achieve certain PFAS-levels several beds in series must be used. To minimize the load (and therefore costs) of GAC/IX, precipitating agents like PerfluorAd can be used. The active ingredient changes the solubility of PFAS. PFASPerfluorAd sludge can be incinerated. The water then is then further treated with GAC/IX (treatment train). Ozonofraction uses the fact that PFAS remain the air-water interface and creates ozone-bubbles which are considerably smaller than regular air-bubbles. Bubbles then can be physically removed. The water then is then further treated with GAC/IX (treatment train). PFAS-contaminated fire run off and cleaning water can also be directly subjected to incineration. More detailed information on the available techniques and associated costs of firewater treatments is available in section E.4.1 of the Annex. 2.7.4.2.7.6. Other risk management targeted at reducing release Industry best practice guidance (e.g. from the Fire Fighting Foam Coalition66) and regulations or guidelines in some EU Member States (e.g. England and Wales67, Bavaria68) already recommend or impose a range of measures to reduce the risk to the environment from the use of fire-fighting foams (see Section 1.3). These cover for instance containment, treatment, 66 https://www.fffc.org/ 67Environmental Protection Handbook for the Fire and Rescue Service, https://www.ukfrs.com/s tes/default/files/201709/Environment%20Agency%20and%20DCLG%20environmental%20handbook.pdf 68 https://www.oecd.org/chem calsafety/portal-perfluorinated-chemicals/countryinformation/germany.htm 53 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS and proper disposal of foams and fire water run-off. However, it is not clear to what extent these practices are being implemented or what their relative effectiveness is. Stakeholder input to the consultation has also highlighted the importance of such measures to reduce emissions of PFAS-based foams, with recommendations made to legally impose retention systems, proof of proper disposal of any contaminated water/liquid, and use of appropriate PPE and cleaning procedures for after-use treatment. A stakeholder also suggested supporting the transition with mandatory fire management plans for every site, which would include a description of the procedure and reasons for the procurement of the specific fire-fighting foams, their storage, use, recovery, containment and treatment. They also proposed setting up centrally managed stocks at specific, well-contained sites in large industrial areas that could be made available to potential users in case of emergencies, in order to control and restrict the use of PFAS-based foams to only the necessary applications during the transition period. This suggestion could help reduce the risk to the environment while allowing a potentially longer period to transition to alternatives, particularly for large industrial sites. In conclusion, a restriction on the use of PFAS in firefighting foams should consider the obligation to apply best practice emission reduction measures during and after the use of PFAS-based fire-fighting foam during the transition periods when PFAS-based foams continue to be used under RMO 1. Therefore, it is proposed that the restriction should require industrial and professional users of firefighting foam concentrates containing PFAS above the defined threshold: a. only use these foams for fires involving flammable liquids ("class B fires"); b. establish a fire management plan for every site and type of foam use which would include a description of the procedure and reasons for the procurement of the specific fire-fighting foam concentrates, their storage, use, containment and treatment and the efforts made to search for PFAS-free alternatives. The fire management plan shall: (i) aim at reducing the use of PFAS foams to the minimum necessary while transitioning to PFAS-free foams, (ii) be reviewed annually and kept available for enforcement authorities; c. ensure the minimisation of emissions of PFAS in the environment and direct and indirect exposure to humans from all activities related to firefighting such as but not limited to: the use of foam concentrates, water-added foams, fire waters, mixtures used to clean the equipment containing PFAS firefighting foams by applying sector best practices. The measures shall ensure: (iv) an adequate protection of workers using PFAS firefighting foams (v) the maximisation of the collection of the waste (vi) that the waste disposal uses methods maximising the mineralisation of the PFAS substances to the extent feasible. For each event of foam use or leakage handling, proof of appropriate management and disposal of the foam concentrates, water-added foams and fire water should be documented and kept available for enforcement authorities. In addition, from six months after entry into force onwards, firefighting foam concentrates containing PFAS above the indicated threshold which are held in stocks and need to be disposed of shall be treated with techniques maximising the mineralisation of the PFAS 54 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS substances. Proof of appropriate disposal should be documented and kept available for enforcement authorities. 3. Assumptions, uncertainties and sensitivities See Annex F. 4. Conclusion Section 1.1.1.7 discussed the need for further regulatory management of the concerns associated with the use of PFAS in fire-fighting foams, based on the following: Significant hazards have been shown at least for some PFAS, including some short-chain PFAS. Many PFAS are highly mobile, highly persistent, and have the potential to accumulate within the environment and living organisms. The continued use of PFAS-based fire-fighting foams and resulting releases to the environment. A lack of existing regulation, and of implementation or proven effectiveness of other risk management measures to address the release of PFAS from the use of PFAS-based fire-fighting foams. A restriction under REACH on the placing on the market only or on the placing on the market and use of PFAS-containing firefighting foams appears to be an appropriate option because: Alternatives are considered feasible for most applications (all except large atmospheric storage tanks), so that PFAS emissions can be eliminated by using fluorine-free products. However, a transitional period would be needed, adapted to each type of use to allow the testing and selection of the most appropriate foam product and the adaptation of the fire extinguishing system, or replacement of parts thereof, including foam tanks and bunds; Other risk management measures that could reduce release of PFAS to the environment are available and are to some extent already being applied; however, these appear unlikely to eliminate the emissions of PFAS from the use of fire-fighting foams as effectively. It appears advisable to address the concern at EU-level, because there is no indication that Member State measures will be forthcoming, and any potential discrepancies in national-level management could have implications for the degree to which the environment is protected and for the functioning of the internal market for fire-fighting foam products. Firefighting foams are traded over the borders and it would not be meaningful or possible to restrict them nationally due to internal market considerations. Furthermore, due to their high mobility and persistence as well as their ubiquity (at least of some PFAS), it appears very likely that PFAS emissions could lead to cross-border pollution. Section 2 assessed the potential conditions of a restriction, in terms of whether it would ban only the placing on the market of PFAS-based fire-fighting foams, or both the placing on the market and the use of those foams. Potential variations across different user sectors, transition periods, concentration thresholds, and potential combination with other risk management measures are also relevant. Two main options have been considered: 55 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS A !pan on the placing on the market }could allow continued use of existing - stocks of PFAS-ba - ing foams, which have been estimated at between 210,000 and 435, tonnes. PFAS emissions related to their use could continue, and this may last for some 10-30 years after the entry into force of the restriction, based on the shelf-life of fire-fighting foams. When stocks are depleted, users would need to buy alternative foams incurring additional costs (compared to the baseline) of around 27m per year in the EU due to potentially higher volumes of alternative foams needed to achieve the desired performance. Before that, installations would need to be cleaned or replaced at potentially significant one-off cost (cleaning could potentially be in the order of up to 1 billion). This would be at least partly off-set by savings, e.g. from lower disposal cost of fire-water run-off (total difficult to quantify) and fluorine-free foams when they reach their expiry date (potentially 100,000s to millions per year), and from reduced clean-up (potentially up to 10s of millions) remediation costs (potentially up to billions over a long time span). However, more information on the total number of sites, real-world use of PFAS per site as well as implementation and effectiveness of best practices in terms of containment and immediate clean-up would be required to assess the extent to which remediation and clean-up could be avoided by using fluorine-free fire-fighting foams. More details on uncertainties, ranges and other potential impacts are presented in Annex F. A ban cm the placing on the market and the use of PFAS-based fire-fighting foams would immediately stop the emissions from the use of PFAS-based firefighting foams. This increased effectiveness needs to be weighed against the additional sociio-econamic implications. The existing stocks of PFAS-based firefighting foams would need to be disposed of (incineration costs estimated at 320 million) and new stocks would need be purchased (subject to replacement costs minus the value of existing stocks already depreciated estimated at around 1 billion). Furthermore, roll-out by suppliers and training/familiarisation would need to be done in a much more compressed timescale, but any potential savings from using alternatives (as discussed above) would also be incurred more quickly. ban on the manufacture, the placing on the market and the use other RMOs (tbdi - Commented EFID42]: Figures to be updated and refined, f Commented P11:143]; To be devdopped It should be noted that these estimates are associated with significant uncertainties and ranges have been estimated. There are other potential economic costs and benefits that could not be quantified. Adjusting the potential restriction to minimise this is discussed further below. Although alternatives are generally considered to be technically feasible in most applications (further testing is required for large atmospheric storage tanks requiring more research and testing), there are also potential implications of the performance of alternatives in some cases, including slower fire suppression, and foams being less flexible and less easy to use. These have not been quantified. It should be noted that there was divergence in the stakeholder input about technical feasibility of alternatives. A few stakeholders have voiced concerns over the potentially reduced fire safety, at least in specific applications, and the associated risk of additional health, safety and economic (fire damage) impacts. However, our analysis has concluded that they are not the mast likely outcome and that large atmospheric storage tanks are the main application for which there is still extensive further testing required. However. since fluorine-free foams have characteristics which differ from PEAS-based foams. for each user. testing of the alternative foam product in conjunction with the foam application ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS exist, but note that this is a high-level estimate based on extrapolation from German data and expert judgement, so the Eurofeu estimate is likely more accurate. 3. The REACH restriction proposal for PFHxA78 states based on personal communication with one stakeholder and on data from the German Federal Association for Technical Fire Safety (bvfa), that in Germany roughly 600 000 hand held fire extinguishers containing AFFF are placed on the market per year, so it is possible that in Germany 6 - 12 million and EU-wide 40 - 80 million extinguishers are in use (i.e. in circulation in total rather than on an annual basis). Given the same underlying data source (bvfa) was used and similar results were obtained, it is likely that this is in fact the same estimation as source number 2 above, with slightly different assumptions. Based on the figures above, the following estimates the total tonnage of PFAS-based firefighting foam in fire-extinguishers in circulation, as well as the annual tonnage placed on the market. The Eurofeu position paper quotes 6-9 litres as the typical size of a fire extinguisher. According to TSF (based on bvfa data), the size can range between 2 and 9 litres. Multiplication of 6-9 litres with the estimated 15 million fire extinguishers yields a range of 90-135 million litres (wider range: 30-810 million litres using 2-9 litres and 60-90 million fire extinguishers) of PFAS-based fire-fighting agents used in fire extinguishers. This would be equivalent to about 90,000-135,000 tonnes (wider range 30,000-810,000 tonnes)79 of PFAS-based fire-fighting agents currently present in fire extinguishers in the EU, or ca 3,600-6,750 tonnes (wider range 1,200-35,000 tonnes) of PFAS-based fire-fighting agents sold in fire extinguishers in the EU annually80. According to personal communication with Eurofeu, the PFAS-based fire-fighting agents in fire extinguishers are either foam concentrate already mixed with water, or a capsule of foam concentrate that is mixed with water when the extinguisher is triggered. That means that only a small share of the fire-fighting agent in the extinguisher is PFAS-based foam concentrate, and the concentration of PFAS in the fire-fighting agent is much lower (2-5g per 6-9 litre extinguisher, or 0.020.08%, according to the Eurofeu position paper) than for the foam concentrates discussed above. To make the 3,600-6,750 tonnes per year of PFAS-based firefighting agents in fire extinguishers comparable to sales of PFAS-based foam concentrates by sector (presented in the previous subsection), they need to be converted: Conservatively assuming that foam concentrates account for 10% of the fire-fighting agent for fire extinguishers would imply some 360-675 tonnes of PFAS-based foam concentrates are used annually in fire extinguishers in the EU2881. Lastly, to sense check this result, it is compared to the tonnage of ready-for-use products estimated in the previous sub-section: Based on Eurofeu data, it was estimated that the total annual EU use of PFASbased fire-fighting foams in the EU is at least 14,000 tonnes but it could be up to around 20,000 tonnes. Figure A.1 (also based on Eurofeu data) puts the share of 78 https://echa.europa.eu/registry-of-restriction-intentions/-/dislist/details/0b0236e18323a25d 79 Assuming a density of approximately 1kg/l. 80 Calculated by dividing the total tonnage present by the average lifetime of 20-25 years, as ind cated by TSF. 81 Calculated as: 3,600 tonnes of PFAS-based fire-fighting agents * 10% = 360 tonnes of PFAS-based foam concentrate. Similarly for the higher end of the range 6,750 tonnes * 10% = 675 tonnes. 63 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS ready-for-use products at 1%, so the annual tonnage of ready-for-use products is around 140-200 tonnes82. This is somewhat lower than the estimated 360-675 tonnes of PFAS-based foam concentrates used in fire extinguishers. However, the data appear to be consistent because Eurofeu specified that not all foam fire extinguishers are included in the category "ready-for-use foams". Even if the share of ready-for-use products was higher than suggested by Eurofeu (Figure A.1), the total tonnage across all sectors would not be significantly affected by the addition of a few hundred tonnes of ready-to-use products, as it was only estimated at an accuracy in the order of magnitude of thousands of tonnes in this report. A.2.3.3.Other information on tonnages from the consultation The following additional information on tonnages was provided in the consultation: Additional fire-fighting foam manufacturers (not covered by Eurofeu's internal survey) provided figures for three different products they manufacture where the PFAS Carboxymethyldimethyl-3-[[(3,3,4,4,5,5,6,6,7,7,8,8,8- tridecafluorooctyl)sulphonyl]amino]propylammonium hydroxide (CAS number 34455-29-3) and 6:2 FTS are used (i.e. all three products use both substances combined). The three products are employed in different sectors: o The first is used by the respondents' customers in airport and marine applications. Of this foam, 700,000 litres are manufactured/imported and 200,000 litres are sold in the EU every year. o The second is used in oil and gas, marine, chemistry and municipal fire fighters applications. 450,000 litres of this product are manufactured/imported in the EU and 250,000 litres are sold every year in the EU. o The third product is used in the oil and gas and marine sectors. 250,000 litres of this foam are manufactured/imported and 100,000 litres are sold every year in the EU. o These volumes are additional to the Eurofeu data presented above. The three foams in sum account for 550,000 litres of annual sales in the EU. Assuming a density of approximately 1kg/l, this would be equivalent to about 550 tonnes of foam that can be added to the Eurofeu total (but would already be included in the EU total extrapolated from Eurofeu data). Hover, given the exact sector split is not known, they have not been added to the sector breakdown. One respondent operating in the field of industrial safety, in particular dedicated to technical support and training, stated that they manufacture 5,000 litres per year of a foam containing a C6 fluorine compound, which is used only for training purposes. As above, this is additional to the Eurofeu data, but has not directly been added because the tonnage or density is not known, One respondent operating in the oil and gas sector provided figures for four fire- fighting foams they purchase; two of these contain poly(1,1,2,2-tetrafluoro-1,2- ethanediyl),alpha fluoro-omega-2-(3- ((caboxylatomethyl)dimetylammonoi)propylaminosulfonyl)ethyl, whereas the other two contain different PFAS that have not been specified: 82 Calculated by multiplying the total tonnage of fire-fighting foams (14,000-20,000 tonnes) with the share of ready for use products (1%). 64 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS o The two products containing poly(1,1,2,2-tetrafluoro-1,2-ethanediyl),alpha fluoro-omega-2-(3((caboxylatomethyl)dimetylammonoi)propylaminosulfonyl)ethyl are used in the offshore oilrig and refinery sectors for spills83, accidents and function tests in process plant fires and trainings. They purchase less than 5 tonnes per year of each of these foams and employ less than 5 tonnes in each instance of use. o The third product is used in the offshore oil and refinery sectors in cases of spills, accidents and function tests in alcohol fires. Similarly to the previous, less than 5 tonnes are bought every year and less than 5 tonnes are employed in each instance of use. o A volume between 30 tonnes and 70 tonnes of a fourth product is purchased every year by the respondent, but no other details have been provided regarding the use of this foam. One respondent operating in industrial safety for the oil refineries, chemicals and petrochemicals sectors provided figures for one foam based on the C6 fluorine compound, which is used for training exercises on large hydrocarbon fires. They purchase 5 tonnes per year of this product and typically employ it 100 days a year. Another respondent operating in the oil refineries, chemicals and petrochemicals sectors provided figures for one product they purchase, which can be used for almost all class B fires. They purchase between 20 and 60 tonnes per year of this foam and in 75% of cases, fires are extinguished with less than 400 litres of foam concentrate. Respondents quoted prices for PFAS-based fire-fighting foams in the range from 2 to 30 per litre for concentrates. For those PFAS based fire-fighting foams for which data on tonnage and price is available, the weighted average price is around 3 per litre, but note that these products reflect only a small share of the total market, so this estimate is uncertain. Some consultation responses suggest that generally speaking, foams providing a higher performance often contain a higher concentration of PFAS which is associated with a higher cost. A.2.3.4. Number of sites using fire-fighting foams No specific data on the number of sites using fire-fighting foams (PFAS-based or fluorinefree) was available. However, in order to estimate the order of magnitude of user sites, the total number of sites in some of the of the main user sectors can be considered: Chemicals/petrochemicals: There are over 10,000 establishments covered under the EU's Seveso III Directive84. One of the main accident scenarios linked to most Seveso-regulated substances is related to fires. Many other facilities with flammable fuels and chemicals below the Seveso Directive thresholds will also require firefighting equipment. 83 AFFF are in some cases also used as prevent on in spills that have not (yet) caught fire. See for instance: https://www.nrl.navy.mil/accomplishments/materials/aqueous-film-foam 84 Analysis and summary of Member States' reports on the implementation of Directive 96/82/EC on the control of major acc dent hazards involving dangerous substances, Final report, https://op.europa.eu/en/publ cation-detail//publicat on/26c9aa63-523e-11e7-a5ca-01aa75ed71a1. 65 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS blanket formed enables the sealing of the flammable liquid surface, impeding the release of flammable gasses with the ambient air. PFAS surfactants in the firefighting foams also prevent the emulsification of the hydrocarbon liquid with the foam, even for water- miscible hydrocarbon liquids such as alcohols, avoiding therefore the risk of fuel pick-up, which would alter the foam structure and make it flammable. AFFF have been considered very effective and also very forgiving with respect to application, proportioning and foam expansion. They allow the use of specific techniques such as sub-surface injection in non-water miscible flammable liquids, application at very low expansion ratios (e.g. delivered by sprinklers, hollow-jet nozzles, non-aspirated hand lines and monitors) or forceful applications. Overall, AFFF made firefighting foams easy to use at a very high level of reliability and performance (Eurofeu 201992) This is a These are particularly relevant features that enables applications in industrial fires for example tank fires, where large quantities of flammable liquid are stored. They are used for training purposes and in a variety of fire incidents, from small fires to the above-mentioned large tank fires, and can be applied both with mobile and semi stationary equipment. According to Eurofeu (Eurofeu, 2019) AFFF are still amongst the first-choice agents for scenarios where the foam needs to be applied over certain distance (vertical and/or horizontal) onto liquid fuel having a certain depth (like large tank farms of flammable liquids). On fires of shallow fuel spills, emulsification does not play a major role in overall fire performance of the foam agent because there is not enough fuel depth for the foam to sink in. These fires (e.g. damaged cars or even road tankers) likely may not require an AFFF. Similarly, municipal firefighting, fires of solid combustibles (so called "Class A-fires") and fires of melting fuels (solid materials becoming liquid due to heat, such as plastics, fats and waxes) do not require AFFF, the latter fire risks being able to be addressed with modern high performing fluorine free foam agents. A.2.4. Fluorine-free alternatives A.2.4.1. Sales of fire-fighting foams by user sector Consultation with Eurofeu provided figures on the yearly consumption of fluorine-free firefighting foams in various sectors in Europe, based on a 3-year average (2016-2018), highlighting a total use of 6,553 tonnes per year. Of these 6,553 tonnes, 2,134 are utilised in fixed systems and 4,418 in mobile systems93. The split by sector is detailed in Figure A.2 below. Notably, it varies considerably from that of PFAS-based foams, with a much larger share used by municipal fire brigades but a much smaller share in the chemical/petrochemical sectors. 92 Eurofeu, 2019 - Pos t on Paper on Fluorine containing firefighting Foams - Responsible Use of AFFF and the likes 93 The number of companies that provided a response on whether the foams are used in fixed or mobile systems is lower than those that prov ded a response for the sectoral overview, therefore in the original data the total tonnage of the former is lower than the latter. To fill this gap, the tonnages for both fixed and mobile systems have been inflated so that their total matches the total in the sectoral split. The original values are 1,259 tonnes for fixed systems and 2,605 tonnes for mobile systems (total 3,864 tonnes). 67 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS o The second is used by the respondent for testing of proportioning systems (i.e. not for fire-fighting), typically 4-6 times per year, with 1,000-6,000 kg used in each instance. They purchase 10,000 kg of this product per year. o The third was due to start testing in autumn 2019, therefore they did not yet have any experience on real fires with this foam. It is expected that this product will be used about 50 times per year, with 1-400 kg used in each instance. One respondent operating in the field of industrial safety, particularly dedicated to technical support and training, provided figures for two different fluorine-free foams, both used for training purposes: o The first (a product shown to contain detergents according to the substance identification task) is used by the respondent for hydrocarbon fires in the oil and gas sector, with a typical frequency of 150 days per year. They purchase 4,000 kg of this product per year. o The second (chemical group of alternatives unknown) is used by the respondent for alcohol fires, about 30 days a year. They purchase 1,000 kg of this foam per year. One respondent providing training in the safety sector gave figures for one type of fluorine-free foam (a product shown to contain detergents according to the substance identification task). This is used only for training purposes on fires of different sizes and in various sectors, such as airports, oil and gas and marine. They purchase 1,200 kg of this product a year and typically use it around 4 hours per week, depending on the training activity. One respondent active in the airport sector provided figures for one fluorine-free foam (a product shown to contain hydrocarbon surfactants and detergents according to the substance identification task), which is used for all aircraft applications and training activities. They purchase 3,600 litres of this foam a year. Approximately 300 litres are used each month, with a typical use of 15 minutes per month. Another respondent working in the airport sector stated that they purchase 5,000 litres per year of a fluorine-free foam (chemical group of alternatives unknown), which is used only for training and system testing. Additional respondents have stated they use fluorine-free foams based on hydrocarbon surfactants and detergents in aviation, offshore oil installations and onshore terminals and refineries, without specifying quantities. Respondents quoted prices for fluorine-free foams ranging from 0.7 to 10 per litre. For those fluorine-free fire-fighting foams for which data on tonnage and price is available, the weighted average price is around 3 per litre, but note that these products reflect only a small share of the total market, so this estimate is uncertain. Although the range is lower and the average is similar to prices of PFAS-based foams (see above), some respondents suggested that fluorine-free foams are around 50% more expensive than comparable foams containing fluorine. However, fluorine-free foams are still predicted to have a growing presence on the market, due to increasing regulations/controls on fire-fighting training and testing. A.2.4.3. Conclusions of the market analysis for fluorine-free alternatives Based on information provided by Eurofeu and additional manufacturers, it has been estimated that at least some 7,000 tonnes, but probably around 9,000 tonnes of fluorine-free firefighting foams are sold in the EU annually. 69 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS A breakdown by chemical group of alternatives (based on the grouping established in the substance identification) is not available, but consultation responses suggest that the main alternatives used are based on hydrocarbon surfactants and detergents. The split by sector of use varies considerably from that of PFAS-based foams, with a much larger share used by municipal fire brigades but a much smaller share in the chemical/petrochemical sectors. Prices for fluorine-free foams range from 0.7 to 10 per litre, with the average estimated around 3 per litre (subject to significant uncertainty). A.2.4.4. Functions provided in the foams and types of fires the foams are used for The fluorine-free fire-fighting foams considered in this analysis are specifically those that can potentially be used as alternatives to the PFAS-based foams. As such, they are potentially used in the same applications. The consultation responses specifically indicated that fluorinefree alternatives are currently used for training, process fires, alcohol fires and fuel fires, as well as for testing proportioning systems and are applied both with fixed and mobile equipment. When it comes to the application of the products, no significant differences between fluorine-based and non-fluorine foams have been highlighted from a market perspective, but this is analysed in more detail in the analysis of alternatives (see Section E.2.). The substance identification (Section B.1) identified the following groups of substances that PFAS-free fire-fighting foams are based on: hydrocarbons, siloxanes, protein foams, detergents. All of these groups largely mimic the function of fluoro-surfactants in the PFASbased fire-fighting foams, for instance hydrocarbon foams use hydrocarbon surfactants95, siloxanes are also primarily used in fire-fighting foams to function as surfactants96 and detergents are by definition surfactants. A.3. Uses advised against by the registrants The analysis in this Annex-XV dossier is based on substances that have been identified as being used in fire-fighting foams. No review of registration dossiers for all of the potentially relevant PFAS substances has been undertaken in terms of identifying any specific uses that are advised against by the registrants. 95 See for example: https://www.fomtec.com/fluorine-free/category38.html or https://www.chemguard.com/about-us/documentslibrary/documents/Martin2009ReebokEcoguardpresentation2010-10-11.pdf. 96 See for example: https://www.nfpa.org/-/media/Files/News-and-Research/Resources/ResearchFoundation/Symposia/2016-SUPDET/2016-Papers/SUPDET2016Hetzer.ashx?la=en. 70 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Annex B: Information on hazard and risk B.1. Identity of the substance(s) and physical and chemical properties The objective of this section is to identify the PFAS (including long and short chain, their salts and precursors, intentionally used or as impurities) present in fire-fighting foams, the constituents of the fluorine-free fire-fighting foams and any non-PFAS fluorinated alternatives (if they exist). In the following, the approach is briefly described (this Section). Then, results are discussed in Section B.1, in separate sub-sections first for alternatives to PFAS in fire-fighting foam that are fluorinated (but not based on PFAS), then for completely fluorine-free alternatives, and lastly for PFAS used in fire-fighting foams. The substance identification was based on desktop research covering: Literature research based on: o Scientific peer reviewed literature (pubmed, google scholar); o Reports or other publications by national and regional environmental agencies; and o Reports or other publications by NGOs. Information gathered in the framework of regulations: o REACH (for example RMOAs, Annex XV restriction reports, RAC & SEAC documents of PFAS substances); o Stockholm convention (for example risk management evaluation, AoA reports, technical paper on the identification and assessment of alternatives); and o Basel convention(technical guidelines). Safety Data Sheets ((M)SDS) and any other information of known producers/associations; Environmental and human (bio-)monitoring data and case studies; and Expert knowledge (international experts). In general, all the above-named documents were screened by using the following search terms: fire, foam, fluor and/or alternative. More specifically, in case the documents covered the analysis of alternatives (e.g. documents by REACH, Stockholm and NGOs) the documents were screened using the search terms fire and foam. This strategy was also undertaken in the screening of more general reports, for example those reports that cover PFAS in general. These kinds of reports were mostly published by environmental agencies. In cases where analytical measurements were reported (case studies, (bio-) monitoring and scientific publications) it was made sure, that an unambiguous assignment to the usage of firefighting foam could be made. Only in cases where this was possible, the respective data was extracted. A different strategy was elaborated for (M)SDS, in this case only the term "fluor" was used. More detail about the specific search terms applied and the specific documents screened is provided alongside the results in the following sub-sections. A matrix was created to collect all potentially relevant information from the literature review, but the identified information is summarised in the following. 71 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS B.1.1. Name and other identifiers of the substance(s) B.1.1.1. Substance identification non-PFAS fluorinated alternatives Due to concerns about their toxicity and regulatory pressure, long chain PFAS (such as C8, see definition later in this section) have been widely replaced by (perceived safer) alternative substances starting from the early 2000s. These alternatives include short-chain substances like C6 fluorotelomer based fluorosurfactants97, but also non-fluorinated substances. The knowledge of the chemical identity of these substances is currently very limited. As reflected in the Terms of Reference of this project, it is clear that a variety of PFAS and fluorine freesubstances are used in fire-fighting foams, but it is not certain if there are any non-PFAS but fluorinated substances that have been or are still being used in fire-fighting foams. The distinction between PFAS and non-PFAS fluorinated substances is the following: PFAS are a fully (per) or partly (poly) fluorinated carbon chain that "contain one or more C atoms on which all the hydrogen atoms are substituted (present in the non-fluorinated analogues from which they are notionally derived) by F atoms, in such a manner that they contain the perfluoroalkyl moiety (CnF2n+1-)." (OECD 2018). Non-PFAS fluorinated substances do not exhibit this particular feature of having "one or more C atoms on which all the H- are substituted by Fatoms". An example for this substance group are silicon dioxide molecules which are perfluorinated. These substances might be used in textiles as an alternative to PFAS98. Based on the length of the fluorinated carbon chain, short and long chain PFASs can be distinguished. Long chains refers to: Perfluorocarboxylic acids (PFCAs) with carbon chain lengths C8 and higher, including perfluorooctanoic acid (PFOA); Perfluoroalkane sulfonic acids (PFSAs) with carbon chain lengths C6 and higher, including perfluorohexane sulfonic acid (PFHxS) and perfluorooctane sulfonate (PFOS); and Precursors of these substances that may be produced or present in products. To identify any potential non-PFAS fluorinated substances used in fire-fighting foams, a literature research in pubmed and google scholar was undertaken, using the following search terms: (("substance" OR "chemical" OR "compound")) AND ("fire fighting foam" OR fire-fighting "fire fighting") As of April 2019, the pubmed search returned 53 hits. However, the relevant hits covered only poly- and perfluorinated compounds. The same result has been found using google scholar. SDS/supplier information, monitoring data, EPAs, NGOs, case studies and legislation were also screened for information on non-PFAS fluorinated substances (simultaneously with the screenings for information on the substance identity of PFAS- and fluorine free-chemicals, discussed below). No non-PFAS fluorinated substances could be identified. In conclusion, the analysis suggests that fluorinated non-PFAS alternatives in the area of firefighting foams do not exist. This was confirmed in personal communication by Zhanyun Wang 97 Fluorosurfactants are synthetic organofluorine compounds with multiple fluorine atoms. They can be fluorocarbonbased or polyfluorinated (Lehmler, 2005). 98 https://greensciencepolicy.org/wp-content/uploads/2015/04/Presentation-Stefan-PosnerPFAS-April-2015.pdf 72 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS (ETH Zrich), an international expert on PFAS chemicals. It was also discussed and not disputed at the project workshop. B.1.1.2. Substance identification - FFF (fluorine-free foams) Because of regulatory pressure and consumer preferences for fluorine-free replacements, a lot of producers of PFAS-containing foams have introduced fluorine-free alternatives. Most of the foams are advertised as intended for use on class B hydrocarbon fuel fires such as oil, diesel and aviation fuels as well as class A fires such as wood, paper, textiles etc. As explained above, various information sources have been reviewed in order to identify any relevant alternative to PFAS in fire-fighting foams. Many of these sources did not provide chemical names or/and CAS/EC numbers. In a lot of sources (e.g. from NGOs, ECHA and Stockholm Convention documents), only very general hints on replacement substances or substance groups have been identified. This includes the naming of the following substance groups: 1. Hydrocarbons; 2. Detergents; 3. Siloxanes; and 4. Protein foams. However, more specific information on substances in FFF was identified in SDS and/or supplier information, some reports published by national authorities, and some peer-reviewed publications. Most relevant information was identified in SDS. As an additional source patents were considered using the google patent search. The results were in most cases the same as for the SDS. A report by the Swedish chemicals agency (KEMI) compiles available knowledge about firefighting foams that were available on the Swedish market in 2014, with respect to chemical content, use, handling and disposal99. Scientific peer-reviewed publications by Hetzer et al. highlighted various sugar-based siloxanes (Hetzer, R. et al. ; Hetzer, R. et al. 2014; Hetzer, R. H. und Kmmerlen 2016; Hetzer, R. H. et al. 2015). However, to our knowledge no CAS-numbers are available for these chemical compounds. In the following, the identified substances are presented in more detail. In general, AFFF concentrates are themselves mostly water, with other components such as surfactants, solvents and stabilisers. The lowering of surface tension to allow formation of foam and hence a blanket over the source of fuel, may be accomplished by use of both fluorocarbon and hydrocarbon surfactants. In this context, some of the substances identified in this task are not believed to be direct PFAS- replacements in terms of being a surface active agent100. In the following, only those substances which were identified by their chemical structure as replacements (R) for PFAS are discussed. It is also possible that some of the identified substances may need to be combined with other substances (for example a hydrocarbon in combination with a detergent) in order to fulfil their capacity as a PFAS-replacement. However, it should be noted that their suitability as alternatives to PFAS-based fire-fighting foams is discussed in more detail in the analysis of alternatives (Section E.2.). 99 https://www.kemi.se/global/pm/2015/pm-6-15.pdf 100 Those substances are for example antim crobial agents that are needed for the b olog cal stabil ty of the foam. 73 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS B.1.1.3. Substance identification - PFAS Generally, most information on PFAS in fire-fighting foams was found in the scientific literature. This is partially due to the fact that SDS and supplier information only indicate general terms like "fluorinated surfactant" without naming a CAS number and/or referring to propriety information. Environmental agencies mostly also cite scientific literature, so this information overlaps with substances already identified in the review of scientific literature. This is also true for information from legislation (REACH, Stockholm, Basel Convention). When searched in pubmed and google scholar, the following search terms were used: ("fluorochemical*" OR "per- and polyfluoroalkyl" OR "perfluoroalkyl" OR "polyfluoralkyl" OR "fluorinated" OR "PFAS") AND ("fire fighting" OR "airport" OR "fire") As of April 2019, this search yielded 86 hits. Those publications were mostly highly relevant, and the substance details were extracted into excel sheets relevant for the next working steps. An additional source of information is case studies and monitoring activities. However, these are considered to be of less importance because mostly only a very limited variety of PFAS substances was covered. Additionally, when environmental/human samples are considered, for fluorinated foams, also environmental and biological degradation processes need to be considered. For example, it is known that perfluorosulfonamides, undergo abiotic degradation as well as in vivo and in vitro biotransformation (DanEPA 2015). With regards to the substances identified in the scientific literature, for a large share it was not possible to find a CAS/EC number. Sixty-three substances were identified by CAS/EC number, while around 213 were only identified by substance name/structure. This lack of CAS numbers may be due to the fact that those substances have been described for the first time by the respective author or are perhaps polymeric substances that do not necessarily have CAS numbers. In general, these numbers might also indicate that a lot of substances have been used that are currently poorly known. The following information relates only to those substances that were fully identified in terms of CAS/EC, substance name and/or acronym. Based on the CAS-identified PFAS-substances that were/are used in AFFF the following grouping is possible, indicated in brackets is the number of CAS-identified substances: Unsubstituted long chain PFAS (14); Unsubstituted short chain PFAS (8); Substituted short and long chain PFAS (12); Fluorotelomers (22); and Others (7). Long Chain PFAS The group of long chain PFAS (defined by OECD as perfluorosulfonic acids (PFSAs) with C6 and perfluorocarboxylic acids (PFCAs) with C8) encompasses the following substances: PFSAs with C6 PFHxS (C6); PFHpS (C7); PFOS (C8); 83 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure B.3 Hierarchical clustering of the identified short-, long-chain and substituted PFAs substances Fluorotelomers Fluorotelomers are defined by having an additional non-fluorinated spacer between the perfluorinated alkyl chain and the charged head group (denotated as number of perfluorinated carbons: number of non-fluorinated carbons). The fully identified substances (i.e. by CAS/EC number) are shown in Table B.9. As shown in the table below, the 22 identified fluorotelomers cover a wide range of positively/negatively charged head groups or combinations of those. Most of the fully identified substances, exhibit the xx:2 structure, where two non-fluorinated carbon atoms are inserted between the perfluorinated carbon chain and the head group. However, in the case of fluorotelomer betaines also xx:1:2 and xx:3 are found. In the latter case, three non-fluorinated carbon atoms are inserted between the perfluorinated carbon chain and the head group. In the case of the xx:1:2 substances, an additional fluorinated carbon is inserted between the perfluorinated alkyl chain and the non-fluorinated spacer. Based on the manufacturing dates that are cited in the respective publications, it can be assumed that the use of fluorotelomers in fire-fighting foams began later than the use of traditional PFAS substances without a non-fluorinated spacer. The following head groups have been identified: Alkylbetaine (AB); Betaine (B); 91 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure B.4 Hierarchical clustering of identified fluorotelomers Other PFAS substances In some cases, perfluorinated substances that do not belong to any of the named groups (long/short-chain PFAS, fluorotelomers, and derivates of PFAS) were identified (Others). These substances are shown in the table below. Also shown below is the substance Dodecafluoro-2methylpentan-3-one, a fluorinated ketone. 96 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS https://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=ENV-JMMONO(2018)7&doclanguage=en Wang, D. Z., Goldenman, G., Tugran, T., McNeil, A., & Jones, M. (2020). Per- and polyfluoroalkylether substances: identity, production and use. Retrieved from Copenhagen: https://doi.org/10.6027/NA2020-901. Available at: http://urn.kb.se/resolve?urn=urn:nbn:se:norden:org:diva-5872. B.1.2. Composition of the substance(s) For each constituent/ impurity/ additive (particularly for those which influence the outcome of the dossier) the following information is provided: Chemical Name EC Number CAS Number IUPAC Name Molecular Formula Structural Formula Molecular Weight Typical proportion % Real proportion (range) in % Chemical name, EC and CAS numbers are provided in the previous section. However, the other pieces of information have not been collected for the present study (with the exception of some example molecular structures provided in the previous section). This reflects that a large number of relevant substances has been identified and that the information is not considered to affect the outcome of the dossier, which proposes a restriction not of the individual substances but rather for the use of PFAS in fire-fighting foams as a group based on the definition provided in Section 1.1.1.7. B.1.3. Physicochemical properties Physicochemical properties of the substances have not been reviewed in detail here, for the same reasons as highlighted under `composition' above. This may need to be considered further when any full Annex XV dossier is drawn up. B.1.4. Justification for grouping See Section 1.1.2. B.2. Manufacture and uses (summary) The table below summarises some of the key results that have been discussed in more detail above (Annex A). 99 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS not available in the registration dossier)109. In addition, the substance with CAS 34455-29-3 also has the lowest PNEC for soil, meaning that, at concentrations higher than 1,33 g/kg (ppb) a risk cannot be excluded. The combination of this value and with its relatively low PNEC for freshwater (0.0326 mg/l), shows, that this substance exhibits more hazard to the environment than any of the non-fluorinated substances. This finding is also supported by the fact that the treatment at WWTPs is ineffective (as shown in the previous subchapter). In terms of partitioning the fluorinated substance CAS 34455-29-3 has a log koc of 1.5, suggesting strong partitioning to treated effluent within WWTPs and release to surface water. Use during live incidents is assumed to be released equally to surface water and soil. This may suggest that the bigger impact for soils would come from live incidents. However, some of the alternatives have both relatively low PNECs and relatively high biodegradation and/or bioaccumulation data. This is true for two alcohols (1-dodecanol and 1tetradecanol). However, in comparison to the two fluorinated substances listed in the table above, both of the non-fluorinated substances are readily biodegradable due to the rapid metabolism of long-chain fatty alcohols in fish, mammals and microorganisms (based on information taken from their registration dossiers). That means that, even if the substance is emitted to the environment in the context of a release from WWTPs or live incidents, it will be biodegraded rapidly. Furthermore, as highlighted in the previous section, based on these properties it could be expected that waste water treatment plants would have a high level of efficacy for the destruction of these substances. Taken together, this review of hazards based on PNECs and data on biodegradation and bioaccumulation shows, that the two fluorinated substances should be considered of higher priority compared to the non-fluorinated substances when it comes to hazards and potential risks to the environment. This is due to the PFAS being both non-biodegradable and having relatively low PNECs for water and soil. Some of the alternative substances exhibit low PNECs, however, this needs to be considered in the context of their ready biodegradation. Further discussion on the hazards of the shortlisted alternatives can be found in Section E.2. B.5.1. Toxicokinetics (absorption, metabolism, distribution and elimination) Not assessed in this study. B.5.2. Acute toxicity Not assessed in this study. B.5.3. Irritation Not assessed in this study. 109 https://echa.europa.eu/de/registration-dossier/-/registered-dossier/17549/1 105 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS B.5.4. Corrosivity Not assessed in this study. B.5.5. Sensitisation Not assessed in this study. B.5.6. Repeated dosed toxicity Not assessed in this study. B.5.7. Mutagenicity Not assessed in this study. B.5.8. Carcinogenicity Not assessed in this study. B.5.9. Toxicity for reproduction Not assessed in this study. B.5.10. Other effects Add text B.5.11. Derivation of DNEL(s)/DMEL(s) B.6. Human health hazard assessment of physicochemical properties Not assessed in detail, see explanation in B.5. B.7. Environmental hazard assessment Not assessed in detail, see explanation in B.5. B.8. PBT and vPvB assessment Not assessed in detail, see explanation in B.5. B.9. Exposure assessment B.9.1. General discussion on releases and exposure An exposure assessment was not in the scope of this study. However, Section B.9 presents estimates of the emissions of PFAS and of the constituents of the alternative fluorine-free firefighting foams to the environment, broken down by environmental compartment (Section B.9.3) 106 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The model has been designed to allow calculation of both quantities of fire-fighting foams manufactured within the European Union, and quantities of finalised fire-fighting foam concentrate imported and used in the EU. Only quantities manufactured within the European Union are assumed to lead to emissions and exposure at the formulation stage. The PFOA Annex XV dossier assumes default worst case emission rates of 2.5% w/w to air, 2% w/w to water (assumed to be waste water system rather than direct release) and 0.2% to soil as a direct release from spillages / deposition during manufacture. In the absence of better data, the same release rates have been applied to the non-fluorinated alternatives. Storage Following manufacture and sale, the fire-fighting foam concentrates will pass into the storage phase of the life-cycle. A proportion of the annual sales will also go directly into use (see in-use phase), with the remainder held in store, sometimes for several years. Data on leakage rates / spillages was not identified during the study, and therefore a value based on expert judgement of 1% of total stocks has been applied. In-use phase The "in-use" phase of the model was then further refined to incorporate different kinds of use and application and how these may affect the type of emission and usage rate (i.e. use at airports vs municipal fire brigades for example). This included data from Eurofeu (see Annex A) on industry sector splits, and usage information from several stakeholders. In their study, Wood (2020) used annual usage rate data from Brooke et al (2004)114 which highlighted that most of the fire-fighting foam in the private sector is used for training (93% w/w). In the absence of better data, it was assumed that, for public fire brigades, use will predominantly be focused on live incidents with a smaller quantity used for training, assumed to be 93% on live incidents and 7% on training. Wood (2000) used for annual foam use rate Ddata from BiPRO (2010)115 and Buser et al (2009)116 which quote usage rates of between 15% and 20% annually117. However, several industry stakeholders provided comments on the PFHxA restriction proposal118 and during the 114 Brooke et al (2004), "Environmental risk evaluation report: Perfluorooctanesulphonate (PFOS), Report produced for the England and Wales Environment Agency. 115 BiPRO (2010), "Study on waste related issues of newly listed POPs and candidate POPs", Commiss on report under framework ENV.G.4/FRA/2007/0066. 116 Buser et al (2009), `Substance flow analysis of PFOS and PFOA in Switzerland. Environmental Studies 0922. Federal Off ce for the Environment, Bern. 117 BiPRO 2010 base their estimates on usage rates against a survey of UK fire author ties completed by RPA in 2004. This suggested that annually 15% of total stocks are used across all sectors (public and private). Usage rates by mun cipal fire and rescue services were higher at between 40% and 50%. Buser et al 2009 base their estimates on remaining stocks of PFOS w thin all sectors (publ c and private serv ces) using 20% of all stocks annually. To maintain a steady flow of business it is assumed that both publ c and private brigades will replace stocks as use occurs, so replacement foam would be purchased annually. 118 See https://echa.europa.eu/restrictions-under-consideration/-/substance-rev/25419/term 109 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS preparation of the present restriction proposal on firefighting foams that these figures are outdated. Instead, they provided alternative estimates. Eurofeu clarified (Eurofeu consultation, February 2021) that in 2004 no restriction on the use of AFFF were in place and fire trainings were not as restricted as today hence were very common and broadly done by all fire brigades. According to Eurofeu, since the first restriction of PFOS the situation changed, foam is only used if ultimately necessary, municipal brigades mostly dealing with a fire incident first with water and only lately with foam, if needed. Trainings on live fires are extremely rare, very expensive and only possible in specific places, most of which in the meantime do not allow using fluorine containing foams. So-called "cold" foam trainings (no fires, foam application training only) are done in the vast majority of cases using "training foams" which are fluorine-free. Eurofeu reported that users from the petrochemical industry indicate that less than 10% of the foam they have on stock is being used in incidents (fires, precautionary foam application and unintended releases from malfunctions) (Eurofeu consultation, February 2021). The Firefighting Foam Coalition (FFFC) indicates in their comment to the PFHxA restriction proposal119 that a 2011 update of the AFFF inventory report estimated that annual AFFF use in the United States for the years 2004-2011 was 8% of the total stock120. As foam uses and practices are similar between the US and EU, FFFC believes that a foam use rate of about 8-10% is a reasonable estimate for the EU during this time period. FFFC however adds that "if 8-10% of foam stocks were being used each year during a period before best practices were implemented or required, use rates for the next 5 years would be expected to be much smaller, maybe as low as 3-5%". WFVD, the German Industrial Fire-Fighters Association carried out a survey in summer 2020 on the use of fire-fighting foam by industrial fire services in Germany and submitted the report during the consultation of the PFHxA restriction proposal121. The survey had a response rate of 12.3% with a total of 96 responses having been included in the report. Remarkably, eight of the ten largest German airport fire services and ten of twelve oil refinery fire services responded to the survey. The share of respondents per sectors is indicated in the below figure. 119 Comment 3010 in Part 2 document available at: https://echa.europa.eu/restrictions-underconsideration/-/substance-rev/25419/term 120 https://www.informea.org/sites/default/files/imported-documents/UNEP-POPS-POPRC13FU-SUBMPFOA-FFFC-3-20180112.En.pdf 121 WFD comment 3158 to the PFHxA restriction proposal - available in Part 5 document at https://echa.europa.eu/restrictions-under-consideration/-/substance-rev/25419/term 110 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The survey results indicate that the industrial fire services of the refineries have by far the largest average amount of foam, both fluorine-free and PFAS-based foam. This is followed by the industrial fire services of airports and the chemical industry. 111 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The survey shows that most fire services are providing PFAS-based as well as fluorinefree foam concentrate. 47% of the responding industrial fire services have both types of foam concentrate available, 36% of the industrial fire services only use fluorine-free foam and 17% only use PFAS-based foam. 112 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The use of foam concentrates in German industrial fire services 113 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS WFD indicates that the results for the average amount of PFAS-based and fluorine-free foam used per year and sector show that foam is used relatively rarely. Especially when the numbers are put in relation to the average stock, it becomes clear that the use of foam is not commonplace, even at industrial fire services. However, WFD highlights that it must also be considered that in a major incident, very large quantities can be used at once - a circumstance that could not be adequately taken into account in this survey and through the presentation of average values (the question referred to the last five years as the observation period for specifying the average value). 114 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS However, from the stocks of PFAS-based foam and annual use indicated above, average annual use rates per sector can be derived and are below 12%: Sector of use Stock of PFAS foam Annual use (m/y) (m) Chemical industry 20 1.4 Automotive 5 0.5 manufacturers Metal processing 14 0.2 industry Airports 36 0.5 Oil refineries 153 3.5 Other industries 7 0.8 Annual use compared to (%) 7.0 10.0 rate, stock 1.4 1.4 2.3 11.4 How long will the transition take? The chart of the mean estimated time it takes to transition to fluorine-free foam concentrate shows that more time is required, especially for the refineries. The other industries all indicate a similar duration of about 2 years for the changeover. However, it should be taken into account that the question was about the duration of the project, 115 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS i.e. from the time of the decision to change. Therefore, the values provided below should be interpreted in relation of each other and not in absolute values. Training and tests Training of emergency personnel and tests of extinguishing systems, vehicles or other mobile equipment hardly ever take place with PFAS-based foam concentrate Those industrial fire services that are still practicing with PFAS-based foam collect the contaminated extinguishing water and dispose of it properly 116 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 117 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Based on the market analysis and stakeholder engagement, the most commonly named products in use were given priority and screened in for the final selection. Based on analysis of the "screened in" set, the highest priority products (most commonly named five) were passed into the next phase. This included: Respondol ATF 3/6 - manufactured by Angus Fire. Moussol FF 3x6 - manufactured by Orchidex Bluefoam - manufactured by Orchidee Re-healing foam RF11% - manufactured by Solberg Re-healing foam RF3x6 ATC - manufactured by Solberg Step 3 - Final selection of substances Once the prioritised set of products was identified, the composition of products was identified (using safety data sheets) and hazard classification based on CLP. Using this approach those substances with hazard classifications relating to human or environmental toxicity were selected for use in the source-flow model. The table below provides details of the specific substances where emission estimates have been developed. Note where ranges have been provided the upper limit has been used for the calculations as a conservative estimate. 122 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Product Re-Healing Foam RF3x6 ATC Re-Healing Foam RF3x6 ATC Substance Starch Cocamidopropyl hydroxysultaine Category Hydrocarbon Detergent CAS number 9005-25-8 68139-30- 0 Concentration in product % w/w >1% <2.5% Hazard classification Not classified Eye Irr tant. 2 Degradation and fate* (E) B odegradable ECHA DB: Will biodegrade in water, 71% degraded after 28 days at 20 Celsius. * All degradation and fate data is based on Pubchem (https://pubchem.ncbi.nlm.nih.gov/), the ECHA database of REACH registered substances (ECHA DB) (https://echa.europa.eu/information-on-chemicals/registered-substances), or in cases where no information was found had been based upon expert judgement (E) 126 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS An assessment of potential treatment scenarios developed in the study128 underlying this dossier suggests that for large infrastructure installations (e.g. airports, petrochemical facilities, and fire-fighter training complexes) the site should be engineered to allow for a 100% capture of materials used in the training activity. Furthermore, for live emergencies at such sites where larger volumes may be used and are expected to be handled, capture of firefighting water should be done as soon as practicable and safe. However, also note that, for live incidents, the releases of firefighting foams are very situation-specific and sitespecific, and, in reality, it may not be possible to retain all runoff from fire-fighting. The specific kind of engineered options (hard surfaces, bunded areas, on-site drainage systems, etc.) will vary from site to site and the specific kind of operation being undertaken. As a further example of the practical application of how a given site may be managed, the UNECE good practice guidelines provide some further insight129: "There are several possible types of systems for the retention of contaminated firefighting water The systems can be installed permanently (i e pre-installed water barriers or permanent retention basins, if necessary with pumping installations) or be provided as mobile facilities (i e fire-fighting water barriers, hoods and sealing pads, mobile storage tanks) " Firewater run-off can then be pumped into tanks and transported e.g. by trucks to treatment facilities. There are several short case studies of fire incidents with a description of retention and disposal of fire-water in Annex 1 of the UNECE good practice guidelines. One further consideration is the management of fire-fighting foam or firewater runoff at either on-site waste water treatment works or municipal waste water treatment plants. Again, this is likely to vary from site to site and is determined in part by the frequency of training and quantities of material that need to be managed. On-site treatment plants would incur a significant cost in the construction and operational phases, as well as requiring a minimum level of throughput to make operations practical. In some cases (e.g. petrochemical works) it is possible that sites already have on-site WWTPs for other purposes and are able to manage firewater runoff as and when needed. In other cases where training is less frequent (e.g. only quarterly / twice a year) use of municipal waste water treatment plants under environmental permitting is more likely. However, also note that where firewater runoff enters drains and is sent to municipal waste water treatment plants, the environmental permits may require some pre-treatment steps. For example, these could include the use of sediment traps to remove solids, an oil/water separator and possibly a granular activated carbon filter before discharge. As a conclusion a distinction needs to be drawn between uses for training purposes and uses for live incidents, noting the potential for greater control over runoff from training compared to live incidents. A review of the evidence suggests that at national level there are regulations in place in several countries over the design and management of fire-fighting runoff for training, and best practice guidelines for live incidents. However, further data on how 128 See Section 6 in: Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in firefighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 129 https://www.unece.org/fileadmin/DAM/env/documents/2017/TEIA/JEG MTGS/UNECE Safety Guidelines and Goo d Pract ces for Fire-water Retent on 14 Nov 2017 clean.pdf 129 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Annex C: Justification for action on a Union-wide basis See Section 1.2. 138 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Annex D: Baseline See Section 1.3. 139 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS E.2.2. Identification of potential alternative substances and techniques fulfilling the function A list of the most common alternative fluorine-free products that are widely used in the EU has been generated. These provide a starting point which can be compared to the risk, performance and cost of PFAS-based products. Alternative techniques could be changes in demand for flammable fuels which would reduce the need for AFFFs. Application of e.g. electric aircraft and phase out of hydrocarbon fuels for vehicles would reduce the needs for AFFFs, but are by the authors of this report not considered feasible alternative solutions in the short term. The selection of fluorine-free products for further analysis has been based on the following criteria: Use - The use of the products has been reported by several stakeholders, ensuring that the products analysed are commonly used in the EU as alternatives for PFAS-containing foams; Chemical group - The products represent different chemical groups according to the grouping in the substance identification, i.e. hydrocarbons, detergents, siloxanes and proteins. Some products may contain a combination of substances from these groups; Technical feasibility - The products do actually represent alternatives/replacements for PFAS- containing foams, including in critical situations (with large fires). Technical feasibility also considers the combination of the foam concentrate, the application system and the application rate to establish whether the alternative is a viable replacement. Case studies of critical applications serve as a starting point for successful replacement of PFAScontaining foams with fluorine-free alternatives. Training foams have been excluded as they are already available and widely used for all applications. ; Manufacturers - The products originate from different manufacturers; Availability - The products are known to be on the market in the EU and are available without further R&D delays or costs; and Complementarity - The products cover jointly all major applications of PFAScontaining foams and can be used in different conditions. An initial shortlist with 30 products from 8 manufacturers was presented at the workshop undertaken as part of the underlying study (see Annex G for more details on the workshop), and participants were asked which were the most commonly used and viable. On the basis of the workshop feedback, further review by the study team and responses from stakeholders, a list of products for further analysis was generated. This is shown in Table E.1 along with a justification of why these specific products have been chosen. For each of the manufacturers, one or two products in the product range has been selected for the more detailed assessment. The selection has been based on the available information on the feasibility of using the alternatives with particular emphasis on products demonstrated as viable alternatives to PFAS-containing foams in airports and the petrochemical sector. The information provided in Table E.1 is supplemented with two representative case studies in Section E.2.4. 141 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The remaining products on the shortlist presented at the workshop were from the manufacturers Auxquimia (EE-3 Newtonian Training foam, and Unipol-FF), Fomtec (the Enviro product range) and the 3F Company (Freedol SF). None of the companies have answered the questionnaire and only limited information on the feasibility has been obtained from the stakeholder consultation. These products were not included in the list, but this does not indicate that these products are considered less efficient alternatives to the PFAS-based foams, merely that less information on the feasibility of using these foams was available for the assessment. Seven substances have been selected in order to strike a balance between ensuring variety in coverage of alternatives and depth of analysis that is possible. It is important to note that during the substance identification, a group of potential alternative fluorine-free products, the siloxane-based alternatives, were identified. These have not been identified as being widely used and, furthermore, at the stakeholder workshop, concerns were raised by governmental stakeholders in relation to PBT and/or vBvP properties of some siloxanes. They have therefore not been selected from the more detailed analysis. One protein-based product, PROFOAM 806G from the company Gepro has been mentioned to be in use during the stakeholder consultation. However, specific data on users, application or feasibility have not been provided by the stakeholder consultation and the manufacturer and products cannot be identified. Protein-bases foams are marketed by Profoam srl (PROVEX AR 6-6), Angus PFAS based foams (TF 3 and TF90 for training purposes) and Stahmer (Foamousse product range). No information on these products has been provided for the stakeholder consultation but one product from the Foamousse product range has subsequently been added to the example list in the table below. Table E.1 Shortlist of fluorine-free alternative products for assessment Product name Manufacturer Chemical group(s) Current use sector of the product where PFASbased products are currently used Reason for shortlisting Other marketed fluorine- free products from the manufacturer for hydrocarbon fires Respondol ATF 3-6% Angus fire Hydrocarbons and detergents Petrochemicals -processing, storage and transport of hydrocarbons and polar solvents Applicable for all types of flammable liquid fires JetFoam ICAO-C (aviation) JetFoam ICAO-b (aviation) Syndura (aviation, forestry) Re-Healing Foam RF1 1% Solberg Hydrocarbons and detergents Petrochemicals - offshore oil installations and onshore terminals and refineries Widely used detailed information on the feasibil ty of using the substances as alternatives for PFAS- 8 other products in the Re-Healing Foam RF product range 142 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Copenhagen airport, the same fluorine-free foam (Solberg Re-healing foam RF3x6 ATC fluorine-free foam) is used for training and emergency response. Timeline in the shift from PFAS foams to fluorine free-foams In 2003, the airport recognised PFAS in the run-off firewater from the airport's training area and its burn pit. This resulted in restrictions on use of PFAS-containing AFFF and later, in 2006, all training with PFAS foams stopped; In 2008, testing with fluorine-free foams was started. Re-Healing foams from Solberg were identified as suitable alternatives; and In 2009, the airport conducted additional tests required by the ICAO ARRF working group. All tests (ICAO foam test and test according to the US Mil-Spec protocol, including the NFPA 403), were passed by the fluorine-free foam carrying airport crash tenders. The results from the UK CAA/ICAO tests also showed that CAFS (Compressed Air Foam System; application of foam with non-aspirating turret)138 were about 40% more efficient in fire extinction compared to aspirated foams. CAFS with PFAS and PFAS-free foams were both shown to be efficient. The PFAS-free foam was implemented jointly with three new airport crash tenders (specialised firefighting trucks designed for use in aircraft rescue and firefighting at aerodromes) with CAFS on all low-pressure outlets. Challenges Along with the implementation of the new firefighting trucks, the training of the firefighters with the new equipment and foams was a crucial issue and initial testing and training caused additional costs (exact cost estimates are unknown). Also, the different viscosity of the PFAS-free foam caused some initial challenges, which were later solved by the adjustment of equipment; and Some of the old trucks continued to be in use and, even though the tanks were cleaned thoroughly, a contamination of the PFAS-free foam with PFAS occurred initially. Costs of replacement Upon implementation of the new fluorine-free alternative, testing and training required ~5,000 litres foam/year. However, with some modifications to the equipment and training, the volume has now been reduced to 3,000 litres foam/year. Optimal efficiency was found at a 6% foam concentration (ICAO Level C) instead of 3% (ICAO Level B), thus larger foam volumes may still be used in certain situations; Costs incurred in the replacement comprised mainly costs for destruction of PFAScontaining foams and additional training and testing. More specific cost estimates were not available in this case. However, it should be noted that the foam supplier also had an interest in supporting the implementation of the PFAS-free foam and carried out some of the foam testing and covered the additional costs; and The investment in new airport crash tenders (specialised fire engines designed for use in aircraft rescue and firefighting) was not strictly linked to the foam replacement, but the coincident introduction of new trucks and foam was seen as having a cumulative benefit. Benefits 138 The difference between aspirating and non-aspirating equipment is that the aspirating device mixes air in the foam/water solution within the nozzle or foam maker, whereas non-aspirating dev ces do not. Typ cal examples of non-aspirating devices are water/fog nozzles, water spray heads and conventional sprinkler heads (Ansul Techn cal bulletin no. 55, https://www.ansul.com/en/us/DocMedia/F-83115.pdf). 156 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Copenhagen Airport is still working on the remediation of previous pollution from PFAS foams. In 2014, works on clean-up, containment and reconstruction of the fire training area were started and required an initial investment of more than 15 million. Currently, the maintenance of the drainage system around the fire training ground costs more than 1.5 million per year and this expenditure is expected to continue for at least the next 80 years. The biggest benefit of switching to a fluorine-free alternative foam is that rainwater and firewater runoff can be discharged though the normal sewer system to the municipality's waste water treatment, thus avoiding long-term clean-up issues and remediation costs in the future. E.2.4.2. Case 2 Offshore production in Norway139 Foams used Equinor, representing 80% of all production on the Norwegian Continental Shelf and equivalent to 50% of total production for the North Sea, have managed to substitute PFAScontaining foams with PFAS-free foams at almost all installations. The substitution is close to completion for ~40 offshore installations and is ongoing for five onshore facilities (terminals and an oil refinery). Fire-fighting foams at offshore installations are used for multiple applications including training, system testing and emergency response of live events. At most facilities, Re-healing RF1, 1% foam from Solberg is used, while some older facilities use Re-healing RF1 3% foam. For a few installations (where there is risk of methanol fire), alcohol resistant foam was used. The 1% and 3% foam products are used for petroleum fires and were chosen because they are regarded as a drop-in replacement for fluorinated AFFF. For methanol fires specifically, Solberg Re-Healing Foam RF3x6 ATC (alcohol resistant foam) is used. Basically, all foam is used for training and systems testing as emergency responses are seldom (have not occurred since the implementation of the substitution). Environmental discharges may also occur due to accidental spills. The crude oil and products are stored in caverns i.e. underground storage tanks. The typical size is 50,000 - 280,000 m3 for crude oils and 10,000 - 50,000 m3 for products. The caverns are filled up with fluids to prevent them from catching fire. Timeline in the shift from PFAS foams to fluorine free-foams In 2010-2012, development and testing of a 1% fluorine-free firefighting foam was carried out as a collaborative project between Solberg Scandinavian and Equinor (named Statoil at that time). The driver for the replacement was concern of the environmental consequences of PFAS-containing firefighting foam released to the sea; In December 2012, the Re-healing RF1, 1% foam (RF1) was first used successfully on the offshore installation Kvitebjrn; 139 Case descript on based on the following sources: IPEN position paper 2019 (https://ipen.org/sites/default/files/documents/the v1 5 final 18 april.pdf) Personal communicat on w th Lars Ystanes, Equinor, 2019 global pfas problem- 157 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS In 2013, the RF1 foam was technologically approved for use by Statoil after an approval and verification process; In 2014, approval for starting the multi-use transition project was obtained, with the aim of implementing the new foams at all Norwegian operated installations with 1% foam systems; By September 2016, 30 of 31 Equinor assets had successfully implemented use of RF1 foam; and In 2018, Solberg launched a modified 1% RF1, with lower viscosity at low temperatures and with a yellow environmental classification (compared to red classification for RF1)140 called RF1-AG. This product went into operational use in 2018 on all new offshore installations. Challenges During the substitution implementation, several technical issues occurred which had to be resolved using additional testing by Equinor: During full-scale testing with RF1, a break-down of the foam proportioner occurred which was initially linked to corrosion related to the use of the RF1 foam. Further investigation identified another reason for the break-down and it was concluded that RF1 had no influence on the foam proportioners; RF1 has a higher density and viscosity compared to the previously used AFFF. Higher density may be a problem for installations with substandard foam pumps. However, most Equinor installations were able to handle the increased viscosity and density with only minor system adjustments. At one installation, the pumps were not able to handle RF1 and the solution for this installation is still under evaluation; and Initial uncertainties related to the temperature tolerance of the foam have been removed. The products currently used have a freezing tolerance down to -19C and acceptable low viscosity at ambient temperature. Costs of replacement For Equinor, the total costs of substitution of PFAS-containing foams at about 40 offshore installations and five onshore facilities has been estimated to be approximately 7 million. This estimate does not include costs related to R&D, and regulatory approval costs, which were undertaken in this case by the foam supplier (Solberg). At a few facilities, adjustment of equipment was necessary, but usually, the same equipment was used and new equipment (and associated cots) was not necessary. These total headline costs can be broken down further to include the following: The cost for support in the multi-use phase has been estimated at 2,500 working hours in the period from August 2013 to September 2016, corresponding to a total cost of approx. NOK 3.5 million (approx. 360,000). This included activities such as planning of implementation together with the supplier, preparation of information letters, support team, follow up on technical issues, etc; The cost related to replacement of foam in storage ranges from 50,000 to 500,000 for the biggest oil installations, corresponding to tank storages of 20 - 120 m3. In total, 140 Environmental colour marking system in Denmark and Norway of The Harmonised Offshore Chemical Notification Format under the OSPAR Convention 1992 indicating substances that should be considered cand dates for subst tution. "Red" substances may only be used in limited amounts and shall be subst tuted. 158 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Alternatives have successfully replaced the PFAS-containing foams in a number of airports. Based on the stakeholder consultation, three different products from three manufacturers have been reported to have replaced applications of AFFF in airports in Denmark (Copenhagen, Re-healing foam RF3x6 ATC), Germany ("one of the biggest airports", Orchidex BlueFoam 3x3, Sweden (Arlanda and other airports, Moussol 3/6-FF), and the UK (Heathrow, Moussol FF 3x6). The alternatives are used for all applications. According to the IPEN report "Fluorine-free firefighting foams (3F) viable alternatives to fluorinated aqueous film-forming foams", all of the 27 major Australian hub airports have transitioned to fluorine-free firefighting (F3) foams, as have the following major hub airports: Dubai, Dortmund, Stuttgart, London Heathrow, and Manchester, Copenhagen, and Auckland141. A case story from Copenhagen Airport demonstrates that some testing, modification of equipment and training has been required. The entire transition period was 6 years. Investment in new fire trucks took place at the same time, but this was not directly required due to the foam replacement. It has been indicated by stakeholders that some airports voiced concerns over efficacy and changes of equipment, but no specific information has been obtained. The same certification tests apply for all airports in Europe and the successful transition in several airports indicates that it should be possible for others. Some alternatives comply with the highest ratings of N 1568,1A/1A for both Part 3 and 4. One stakeholder noted that high ambient temperatures can influence the performance of foams as demonstrated in an incident in Dubai. However, as mentioned above all 27 major Australian hub airports have transitioned to fluorine-free firefighting (F3) foam indicating that PFAS-free foams are also being applied at high ambient temperatures. One stakeholder (a supplier of AFFF and alternatives) with experience in transition in a German airport states that that experience from a large number of tests done in the past 10 or so years indicates it is possible to change 99.9 % of all current scenarios to PFAS-free products. Upstream petrochemical sector Equinor, the largest operator on the Norwegian continental shelf, has successfully replaced AFFF in about 40 offshore installations and five onshore facilities. At a few facilities, adjustment of equipment was necessary, but usually the same equipment was used and new equipment was not necessary. At one installation, the pumps were not able to handle the alternative. The company had some challenges with the density and viscosity of the alternative foams initially used compared to the traditionally used AFFF, e.g. by lower ambient temperatures. This was solved by modifications of the alternative product. The shift took approximately eight years from the first tests to when the modified alternative was introduced on all installations. Municipal fire brigades and forestry PFAS-free alternatives are readily available for these areas and, as shown in the market analysis, account for more than 60% of the total market. No data on costs of substitution 141 Fluorine-free firefighting foams (3F) viable alternatives to fluorinated aqueous film-forming foams (AFFF), IPEN Stockholm Convention POPRC-14, Rome, September 2018 160 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS o Aspirating monitor142; and o "Non aspirating" monitor with appropriate foam characteristics. Tank fires ~60m+ diameter No reason to doubt results can be extrapolated to >80m +) or bund fires: o Foam pourer. When used at lower rates than NFPA using CAF application: Tank fires ~15m+ diameter (no reason to doubt results can be extrapolated to >25m+): o Monitor application. Tank fires ~80m+ diameter (no reason to doubt results can be extrapolated to >100m +) or bund fires: o Foam pourer. It is stated in the presentation that test results for some conditions are still missing and LASTFIRE is going to work on these issues: specifically, polar solvent tests - foam application from longer distances, other foams/combinations of foam/application methods, tactics for life safety situations and optimising properties. As indicated above, it can be concluded that even in large tanks alternatives can be applied, but the safety margin may be lower than for the PFAS-based foams. According to stakeholders, the largest risks are associated with fires in large tanks of crude oil because of the higher risk of boil-over. One stakeholder mentioned that fires in large tanks of ~40m are however very rare in the EU and they could not identify any such fires in Europe in the last 10 years. A study by the Fire Protection Research Foundation (USA) determined the fire extinguishment and burnback times for five fluorine-free foams (FFF) and one short chain C6 Aqueous Film Forming Foam formulation (AFFF) as a function of application rate and foam discharge density for a range of test parameters including foam quality/aspiration, fuel type, water type and fuel temperature143. In summary, the authors conclude that PFAS-free foams have come a long way but there is still a lot more to learn about their capabilities and limitations. Furthermore, they conclude: "As of today, FFFs are not a "drop in" replacement for AFFF However, some can be made to perform effectively as an AFFF alternative with proper testing and design (i e , with higher application rates/densities) "143 No specific cases with successful 100% transition in installations with large tanks have been identified. According to stakeholders some examples exist where PFAS-free foams are used for the majority of applications but PFAS-based foams are still stored for use in emergency situations with large tank fires. A reported challenge in petrochemical processing and storage tank farms is the presence of tanks with different liquids that may require different alternatives because one alternative cannot be used for all the liquids. One supplier indicated 142 Fire fighting monitors are a controllable high-capacity water jet used for manual or automatic fire fighting 143 Back, G.G., Farley, J.P. (2010). Evaluation of the fire protection effectiveness of fluorine free firefighting foams. Fire Protection Research Foundation. 162 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS that in some instances in the petrochemical industry two different alternatives could be required whereas another manufacturer indicated that even more than two may be required if many different liquids are stored. As reported elsewhere, in the chemical/petrochemical sector approximately 93% of the foam volume is used for training. Most of the manufacturers provide PFAS-free training foams that mimic the AFFF and which are used for training. One manufacturer indicated that the PFASfree training foams were not used in live-fire training ("hot training") As indicated in the market analysis, PFAS-free alternatives account for 19% of the volume used in the chemical/petrochemical sector, but a major part of this is likely to be for training purposes. However, in terms of performance of PFAS-free foams, a study published January 2020 by the US Fire Protection Research Foundation144 concludes after comparative testing of both types of foams that: "The AR-AFFF performed well against all test fuels included in this assessment (IPA, Heptane, and Gasoline (MILSPEC and E10) The F3s did well against heptane but struggled against some of the scenarios conducted with IPA and gasoline (both MILSPEC and E10), especially when the foam was discharged with a lower foam quality/aspiration The F3s required between 2-4 times both the rates and the densities of the AR-AFFF to produce similar results against the IPA fires conducted in with the Type II test configuration During the Type III tests, the F3s required between 3-4 times the extinguishment density145 of the AR-AFFF for the tests conducted with MILSPEC gasoline and between 6-7 times the density of the AR-AFFF for the tests conducted with E10 gasoline From an application rate perspective, the F3s typically required between 1 5 to 3 times the application rates to produce comparable performance as the baseline AFFF for the range of parameters included in this assessment " Referring to this study, Eurofeu indicated146 that this has significant implications for users: Increased storage volume of foam concentrate; Bund areas in tank farms may need to be retrofitted to cover a significantly higher volume of liquids, associated with an emergency response action; Significantly increased firewater retention capacity is required, which can cause reasonable difficulties in retrofitting operating sites to the new demand; Cut-back on tank storage capacity to leave sufficient space for additional firefighting agent application: Tank owners may have to reduce the maximum stored volume in tanks to leave enough space for containment of an increased amount of firefighting foam; Studies of the US Naval Research labs147 clearly indicate that F3 require significantly higher application rates (see above) This indicates that, even though fluorine-free foams seem to overall perform well against different types of fuels, different application rates and methods may be needed to achieve 144 "Evaluation of the fire protection effectiveness of fluorine free firefighting foams", Fire Protection Research Foundation report, January 2020; https://www.nfpa.org/News-and-Research/Data-researchand-tools/Suppression/Evaluation-of-the-fire-protection-effectiveness-of-fluorine-free-firefightingfoams 145 The authors describe the "extinguishment density" as the total amount of foam needed to extinguish a fire of a given size. 146 Comment #2893 to the PFHxA restriction proposal: https://echa.europa.eu/restrictions-underconsideration/-/substance-rev/25419/term 147 "Fuel for Firefighting Foam Evaluations: Gasoline vs Heptane", Naval Research Laboratory, June 2019, https://apps.dtic.mil/dtic/tr/fulltext/u2/1076690.pdf 163 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS the desired effect, resulting in the need to implement several technical adaptations to the overall fire extinguishing system of a site, including foam storage tanks and bunds. Sufficient time is therefore needed to allow users to implement these technical adaptations to transition to PFAS-free alternatives in order to maintain the necessary fire safety levels. According to a chemical company having commented the PFHxA restriction proposal on 26/06/2020, to their knowledge, there has not been any major Class B fire (e. g. in a tank farm of a processing plant in chemical industry) that was extinguished using PFAS-free foams yet148. They further indicated that there has been a very limited number of larger-scale tests made by LASTFIRE since 2018, so there is not enough evidence on transferability of smallscale test results for Class B fires to large-scale Class B fires. This poses the risk that a given Class B fire cannot be controlled using fluorine-free foam and the tank needs to be burnt down with all socioeconomic consequences including loss of reputation which both need to be prevented. The company raised also the following issues: - Fixed and semi-fixed extinguishing systems e. g. for tank farms and tank pits might require a general overhaul when switching to fluorine-free foams. - Fluorine-free foams require significantly higher application rates which cannot easily be achieved in existing installations. This can result in the need to partial or complete exchange of foam extinguishing systems with large capital invest. - Increased fire knock-down times and application rates will also lead to insufficiently dimensioned firewater retention systems, such as diked areas around tank farms or processing plants configured for AFFF-AR149 use. Diked areas usually cannot be heightened in existing structures due to technical constraints. If constructional changes are possible at all, they usually require disproportionate financial efforts. - Increased firewater demand might require infrastructural changes throughout chemical sites that cannot be realized due to disproportionate financial efforts. - If changes need to be made to semi-fixed, fixed and mobile equipment, at least a decade will be needed for planning, budgeting and execution of the resulting projects. - Processing plants and warehouses pose similar risks as tank farms, as they can contain large quantities of flammable liquids potentially spreading to the plant pit or within the warehouse with surface area > 500 m However, other stakeholders of the same sector did not report the need for such large infrastructure overhaul. 148 Comment 3046 - see in "Part 3" document available at https://echa.europa.eu/restrictions-underconsideration/-/substance-rev/25419/term 149 AFFF-AR: Alcohol-resistant aqueous film forming foam 164 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS E.2.5.4. Economic feasibility The available data indicates that the most significant one-off costs to transition to fluorinefree foams are associated with the following: Replacement of foams in storage. For Equinor, the costs of replacement of AFFFs was 5/l corresponding to 5.5 million; Destruction of replaced AFFFs. In addition to costs of about 1/l for the destruction of the replaced AFFFs, corresponding to a total of 1 million; Decontamination of equipment. The available cases do not indicate significant costs of decontamination of equipment. The equipment has typically been drained and decontaminated by cleaning with washing water which was discharged to waste water or surface water. However, the costs of cleaning of equipment will depend on the requirements as to the decontamination level and discharge of cleaning water. According to information from manufacturers, it may in some instances be less expensive to change part of the equipment than to clean it especially for stationary equipment. Stakeholders have reported, the requirements are different between Australia and New Zealand resulting in large differences in the costs of decontamination of equipment (specific data have not been obtained); Management of the transition process. Reported at 0.36 million for Equinor i.e. less than 10% of total transition costs; R&D and regulatory approval costs. These costs are usually covered by the manufacturers of foams and reflected in the price of the alternative foams; Adjustment and replacement of equipment. The available cases indicate that the costs of replacing equipment has been small in comparison to the cost elements listed above. According to stakeholders, extra storage capacity is not always required; and Training in the use of new products. The available cases do not indicate additional training costs; these are covered by the costs of testing and adjustment of equipment. Regarding the effective price of alternatives, three interviewed manufacturers of PFAS-based foams and alternatives consider that the effective price is more or less the same and within +/- 20%. In accordance with this, additional recurrent costs for alternatives used in the aviation sector, stakeholders have reported that the effective price of the alternatives (taking efficiency of alternative into account) is more or less the same as the price of the AFFF used before the transition. The case from the offshore sector reports extra costs varying between +5% and +30% depending on application with total extra costs slightly below +30% as compared with the AFFFs used before. This may reflect the more diverse scenarios in the offshore petroleum sector. The reported shelf lives of alternatives range from >10 years to 20 years. Shelf life of PFASbased foams is reported to be typically between 10 years and 20 years (to a maximum of 30 years)152. In general, the shelf life of the alternatives does not seem to be shorter than the 152 Proposal for a restr ction: Perfluorohexane sulfon c acid (PFHxS), its salts and PFHxS-related substances https://echa.europa.eu/documents/10162/a22da803-0749-81d8-bc6d-ef551fc24e19 166 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS shelf life of PFAS-based foams and no extra costs, as a consequence of differences in shelf life, have been indicated by stakeholders. The main advantages of using alternatives in the aviation sector are that the rainwater and firewater runoff from training grounds can be discharged though the normal sewer system to the municipality's waste water treatment system, thus avoiding long-term clean-up issues and remediation costs in the future. The case from Copenhagen airport demonstrates that works on clean-up, containment and reconstruction of the fire training area were started and required an initial investment of more than 15 million, and currently, the maintenance of the drainage system around the fire training ground costs more than 1.5 million per year and this expenditure is expected to continue for at least the next 80 years. At offshore installations, training foams are typically discharged directly to the sea and it is not considered feasible to avoid this discharge by collecting and treating the AFFF-containing firewater. As indicated above, the costs of destruction of PFAS-based foams is about 1/litre (a more detailed description of destruction costs is provided in section E.4). The costs of destruction of the PFAS-based foams is likely to be incurred in any case when the foams expire (exceed their shelf life). In the past the PFAS-based foams were also used for training which meant that stocks were used before they reached the end of their shelf life. According to information from stakeholders it is today common to store PFAS-based foams which have reached their shelf-life whilst waiting for a less expensive solution for disposal. In a scenario where PFASbased foams are used for emergency situations and PFAS-free foams are used for training, a cost of about 1/litre for destruction of the PFAS-based foams by the end of their service life should be expected. E.2.5.5. Comparison of application area An overview comparison of the use of fluorine-free alternatives in different applications is provided in the table below: 167 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS E.3. Restriction scenario(s) See Section 2.3. E.4. Economic impacts E.4.1. Socio-economic analysis Section 2.3 summarises the main effects (i.e. anticipated responses from the supply chains along with associated impacts) resulting from the two restriction scenarios and identified the ultimate impacts to be assessed. These ultimate impacts are described below. 4.1.1. Cleaning of equipment: costs and remaining contamination A restriction on the placing on the market of PFAS-containing fire-fighting foams (RMO 1) would allow users to continue using their stocks of foams, but once they are depleted, users would be forced to switch to alternative (fluorine-free) foams. A restriction on the use of PFAS-containing fire-fighting foams (RMO 2) would require this switch to happen immediately when the restriction comes into force (or before). During the storage of PFAS-containing foams, fluorinated surfactants settle on the walls of the tanks as well as in pipe and hose lines of fire-fighting equipment. These would leach into any new foams filled into the equipment and therefore contaminate the new fluorine-free foams with PFAS, leading to continued PFAS emissions154. In order to control these emissions, equipment previously used for PFAS foams may be required to meet a minimum concentration limit of remaining PFAS, which can potentially be achieved through cleaning. This sub-section discusses the feasibility of achieving certain remaining concentrations of PFAS through the cleaning of equipment, with a focus on the associated cost. The analysis of alternatives has concluded that currently available cases of transformation to fluorine-free foams do not indicate significant costs of decontamination of equipment (including disposal of the liquid used for cleaning), with relatively simple methods being applied. However, the costs of cleaning of equipment will depend on the contamination thresholds requirements. According to information from manufacturers, it may in some instances be less expensive to change part of the equipment than to clean it, especially for stationary equipment, so this is also discussed below. Wood (2020) identified the following techniques to clean PFAS-containing foam from equipment are: The use of hot water and detergents in a 32-stage legacy foam decontamination process (stakeholder consultation response). This technique is reported to result in all appliances achieving PFAS levels below 1000ppt and one-third of appliances being below 70ppt. An independent body oversees the 154 Bavarian Ministry of the Interior, Sport and Integrat on, and Bavarian Ministry of the Environment and Consumer Protection: Environmentally friendly use of fire-fighting foams. Available at: https://www.bestellen.bayern.de/applicat on/eshop app000000?SID=578672032&ACTIONxSESSxSHOWPIC(BILDx KEY:%27stmuv all 00001%27,BILDxCLASS:%27Artikel%27,BILDxTYPE:%27PDF%27) [In German] 171 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS process and measures the PFAS concentrations achieved. The approximate cost of this process is 12,300155 per appliance.; and For stainless steel tanks, glass fibre reinforced plastic and polyethylene tanks, following the discarding of the foam, tanks are rinsed with hot water (5060C) and then filled again with hot water for at least 24 hours156. This process is repeated three times in both the tank and any foam carrying pipes and fittings, and the water from these rinsing operations passed into the sewage system and treatment plant. This is recommended in some government guidance157. No information could be identified concerning the costs of this technique or the remaining contamination levels achieved. Several stakeholders commented on the feasibility of cleaning techniques to remove PFAScontaining foams from equipment. One stakeholder considered achieving PFAS contamination levels below 100 ppb to be unrealistic in most cases (from the Helsinki September 2020 stakeholder workshop) and one stakeholder considered it to be almost impossible to achieve a contamination level of zero in a one-digit ppb framework with another stakeholder also commenting that the cleaning of systems and equipment is unlikely to bring the level of residual PFAS to zero. One stakeholder that has transitioned to fluorine-free foams (in the petrochemicals sector) reported that they had aimed for and achieved a level of 0.001% (10,000 ppb). To put this into context, the average concentration of PFAS in PFAS-based firefighting foams is some 2-3% (20-30 million ppb). One stakeholder commented that the level of cleanliness achieved by cleaning techniques would vary depending on the equipment and material being cleaned. The need to accommodate an allowance for residual legacy PFAS even after equipment has been cleaned was also discussed. Stakeholders also commented on how cleaning techniques and costs may be impacted by different PFAS contamination thresholds. Where contamination threshold levels are set high, following the cleaning of equipment, a higher level of residual PFAS-containing foam would be allowed to remain (compared to if a lower threshold limit were set). One stakeholder therefore considered the implementation of a high contamination threshold to be "pointless", due to its reduced effectiveness in eliminating PFAS emissions. With a low contamination threshold level, a lower level of residual PFAS-containing foam will be allowed to remain in equipment following cleaning and cleaning will be more costly than if a higher threshold level were set. Also, where contamination levels cannot be achieved through cleaning, equipment will need to be replaced at a cost. Equipment replacement is more likely to occur where threshold levels are set low. There are potentially significant costs associated requirements for cleaning or replacement of equipment, if a low threshold is set for residual PFAS concentrations (following use of the alternatives in the same equipment as PFAS-based products). The market analysis (Annex A) estimated that there are likely to be several tens or potentially hundreds of thousands of facilities with equipment that contains fire-fighting foams. If all of these require extensive cleaning using techniques such as the decontamination process described above (and costing 12,300 per appliance), the costs of cleaning could be in the region of 1 billion (based on an assumed 100,000 appliances needing cleaning). If a less stringent threshold concentration is used, the costs would potentially be significantly lower. 155 Convers on rate of 1 EUR = 1.62470 AUD applied. It is assumed that this cost includes treatment of the waste water generated, although the stakeholder response d d not specify that. 156https://www.bestellen.bayern.de/application/eshop app000000?SID=578672032&ACTIONxSESSxSHOWPIC(BIL DxKEY:%27stmuv all 00001%27,BILDxCLASS:%27Artikel%27,BILDxTYPE:%27PDF%27 [In German]. 157https://www.bestellen.bayern.de/application/eshop app000007?SID=147496132&ACTIONxSESSxSHOWPIC(BIL DxKEY:%27stmuv all 00001%27,BILDxCLASS:%27Artikel%27,BILDxTYPE:%27PDF%27) [In German] 172 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Regulation in Queensland, Australia, allows for threshold concentrations for replacement foam stocks to be 10ppm (mg/l) for PFOA/PFHxS and 50ppm (mg/l) for PFOA158. Additionally, one stakeholder commented that newer C6 foams are purer and have lower concentrations of impurities than older C6 foams and suggested that different threshold levels for different PFAS-containing foams may be required. For confirmation that threshold levels have been achieved, cleaning techniques may need to be professionally endorsed or, following cleaning, the presence and concentration of remaining PFAS tested. Stakeholder responses reported some concern over the suitability of existing methods to measure and detect the presence and concentration of remaining PFAS. One stakeholder reported that measuring very low concentrations e.g. at ppb-concentration was not possible. One stakeholder suggested that following cleaning, an assessment should be undertaken at an accredited laboratory for verification that threshold levels have been achieved. Stakeholder responses suggested that laboratories are able to analyse down to a level of 30-150 ppb. In the REACH restriction on PFOA, a concentration limit of 25 ppb of PFOA including its salts or 1,000 ppb of one or a combination of PFOA-related substances was adopted, based on the capabilities of analytical methods according to the RAC's opinion on the restriction dossier. Information on the cost of analysis was not provided. A cost analysis concerning the measurement of cleaning success could therefore be done as part of this analysis. Where threshold limits cannot be achieved through cleaning techniques or where cleaning techniques are too difficult or too costly to achieve, the replacement of equipment is likely to be required. The cost of replacing equipment will vary across industries and appliances. Table E.1 provides an example of the potential costs for the replacement of fire extinguishers, where cleaning techniques do not succeed in attaining threshold concentration levels, or the cleaning process costs more than the cost of replacement. It is assumed that these costs represent only the replacement cost of the equipment and do not include the replacement cost of equipment plus foam, nor the cost of disposal of the old equipment. Figures for the total number of fire extinguishers existing and currently using PFAS-based foam have been obtained from the Market Analysis (the lower end of the range is based on a Eurofeu position paper and the higher end considered a more uncertain high-level estimate based on extrapolation from German data and expert judgement). The stakeholder consultation also revealed that the cost for a new extra foam tank in a fire truck is 35,000 for a fire brigade providing industrial fire protection. However, information of the number of existing foam tanks containing PFAS fire-fighting foam was not provided and therefore cost analysis for their replacement has not been estimated. Table E.1 Estimated costs for the replacement of fire extinguishers in the whole of the EU 1 per replacement extinguisher 3 per replacement fire extinguisher 5 per replacement fire extinguisher 15 million fire extinguishers to be replaced 15 million 45 million 75 million 90 million fire extinguishers to be replaced 90 million 270 million 450 million Note that these costs do not include the cost of foam disposal from cleaning. Estimated costs of fire extinguishers were obtained from stakeholder consultat on and t is not clear whether the costs of fire extinguisher replacement 158 Obtained from stakeholder consultation. Also available online here: https://www.qld.gov.au/environment/pollution/management/disasters/investigation-pfas/firefighting-foam/policyoverview 173 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Different institutions (manufacturers, remediation companies, public authorities) define guidelines and instructions for the cleaning process, which fire departments and companies can make use of. Based on current literature and stakeholder input, the following passages encompasses current go-to techniques. In some cases, the described techniques are the same as for the disposal of PFAS-contaminated fire run-off or cleaning water described in section X. Here, only information to the cleaning of equipment is highlighted. "Non cleaning" There is no official guideline that lays out the practical details the transition from PFAS-based foam to fluorine free foam, describing for example cleaning procedures and accepted remaining levels. Thus, companies and fire brigades have been developed their own replacement strategy. Based on the input of stakeholder this included, in comparison to "cleaning techniques", no washing steps with water. One stakeholder from Germany shared their experiences after transition from C8-based foam (3M Lightwater which is supposed to be based on PFOS) to C6-based foam without a cleaning procedure. After the replacement, the C6-based foam was tested for its PFAS content and high concentration of PFOS were found. In the end this observation led to the development of a cleaning procedure specialized on foam concentrate tank located at industrial fire brigades. This procedure is explained in detail in section X. Another stakeholder from Norway stated that when a first round of replacement of PFAS to non-fluorine foam took place, no official cleaning protocol has been used. The PFAS-foam was simply drained and new foam (fluorine free) was filled in. However, follow-up measurements then showed that PFAS were still detectable. In terms of replacement procedure, according to one stakeholder, the PFAS-foam was simply drained and new foam (fluorine free) was filled in. No more information available. The remaining PFAS concentrations from legacy C8-contamination levels as measured by the PFOS-concentration were reported to be 28.000 g/kg (which is higher than the threshold of 10,000 g/kg according to the POP-regulation (10 ppm)). The stakeholder from Norway used a limit is 0,001% (10 ppm) PFAS and had to refill tanks twice in a couple of cases to get below this limit. No information on costs of the actual replacement strategy was available. Secondary costs are due to the incineration of the replaced foam. As highlighted above, both stakeholders have been faced with contamination of the new foams with legacy PFAS-substances (like PFOS). Based on this contamination both stakeholders decided to develop cleaning strategies and had to start the process again. Conclusion on available cleaning procedures for firefighting equipment The following conclusions can be made for available cleaning procedures for firefighting equipment: The procedures described in this report were authored by regional (Germany) and national (Australia) authorities, private companies (PerfluorAd & Arcadis), associations and lastly by manufacturers of foams (BioEx). All of them use extensive cleaning steps with (sometimes hot) water, which can also soak in overnight 175 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.1.2. Other options than substitution and their impacts This section discusses what other responses to a restriction than using alternatives are likely (if any), and their socio-economic impact. Theoretically, in response to PFAS-based firefighting foams becoming unavailable, users could respond by eliminating the need for the use of fire-fighting foams. As discussed in the market analysis, the main application for PFASbased fire-fighting foams are class B fires (flammable liquids and gases). Hence, to eliminate the need for fire-fighting foams would (in principle) require stopping the use of flammable liquids or gases, or accepting a situation where fires are less well controlled than at present. While this may be possible in a limited number of specific applications where they are not crucial, it seems unlikely in most cases. The consultation has also not specified any likely other responses than using alternatives. Therefore, no other options are considered. 4.1.3. Fire safety: impacts of technical performance of alternatives Both scenarios 1 and 2 would lead to a transition to alternative foams. The transition associated with RMO 2 would be faster as existing stock would need to be disposed of at the same time. The key socio-economic issue under both scenarios is the likelihood of fires being extinguished effectively and without delay, compared to the situation using PFAS based foams. The key issues in the technical feasibility of alternatives are three -fold. First, do the alternatives effectively put out fires so that life, environment and property are not at additional risk? Second, if so, are there delays in the duration over which the alternatives can address these fires, considering the technical ability to deliver greater volumes of foams to the fire? Third, do the alternatives have relevant and reliable safety standards so that downstream users can purchase and use these alternatives with confidence, making allowance for testing in users' specific systems? This sub-section discusses the difference in the fire safety performance through the use of alternative fire-fighting foams. These effects are quantified where possible, and drawn out qualitatively where not. This section draws directly on the analysis of alternatives (AoA). As in previous sections, whilst the AoA started with a long-list of some 30 alterative foam products, it focussed on a subset of seven judged to be illustrative of the efficacy of these. The evidence below focusses on these specific products but refers to wider evidence were relevant. Table E.2 provides a summary of the key information. Table E.2 Effectiveness of alternatives - summary Alternative Attained performance standards? Informat on from `real world' use Respondol ARF 36% Yes - 2 (EN 1568 Parts 3 and 4) None Identified RE-healing foam RF3X6 ATC Yes x 4 (EN 1568 Parts 1 and 2, and ICAO Levels B and C)) Yes - Copenhagen Airport & Norwegian Offshore oil sector and Melbourne Fire Brigade. Additional stakeholder informat on Can be used for use in `all types of flammable l qu d fires'. Has been in used in Mun cipal Fire Brigade appl cations - both in training and operational fires. 179 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS RE-healing foam RF1-1% Yes - 1 (EN 1568 Part 3) Yes - Norwegian Offshore oil sector. Consultees state this alternative can be used at offshore oil installat ons and onshore terminals and refinery. Moussol FF 3X6 (F15) Yes x 5 (DIN EN 1568: Part 3 (Heptane): IIIB/IIID, Part 1: Medium ex. - Part 2: High ex. Yes -Swedavia, Heathrow Airport (UK), Norwegian Petrochemical sector. Has been in use at Heathrow Airport (UK) since 2012. See case study. ICAO Low expans on foam Level B DIN EN 3 21A). Foam Mousse 3% F-15 Yes (x1) (EN 1568 Part 3 heptane) None identified (but consultation has confirmed this is in use) Consultees state this alternative is largely used in marine appl cations and is only used for smaller fires (unsuitable for aviation, for example). Epocol Premium Yes x 6 and 1 in progress. EN 1568 - 1: Conform EN 1568 - 2: Conform EN 1568 - 3: 1A / 1A EN 1568 - 4: 1A / 1A None identified (but consultation has confirmed this is in use) Manufacturer states this alternative can be used in all sectors: airports, marine, mil tary, chem cals, oil and gas, mun cipal fire fighters and from fixed mobile and CAFs. Hydrocarbon fires, all types of flammable polar solvent liquids Consultees indicated this as a possible substitute for large tank fires, but further testing was necessary. Oil industry: LASTFIRE Forest fire standards: CEREN Certif cate Certif cation in progress : UL 162 / GESIP). Orchidex Blue Foam 3x3 Yes x 4 (EN 1568 Parts 3 and 4, Oil industry: LASTFIRE, ICAO Level B)) Yes - German airport are reported to be using the product. Consultees indicated potential for additional volumes and/or time to suppress fires may occur for some fuel types, but for others, the performance is the same as for PFAS foams. Effectiveness of foams The central finding, based on evidence from the analysis of alternatives, the stakeholder consultation and the workshop is that from a technical standpoint, no stakeholder concluded 180 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS that alternatives are not technically feasible, at least for the majority of uses. However, as indicated above, a transition period is required to adapt and test the alternative firefighting foam with firefighting equipment and the firefighters to get trained with them. This is particularly true for the oil, petrochemical and chemical sector which can face fire scenarios for which the alternatives have not yet been proven to be fully effective in real large fire incidents. As noted in the AoA, in aviation several airports have successfully transitioned, as have Municipal Fire Brigades and companies active in offshore oil and gas operations and the marine sector. Evidence indicates that one segment - liquid fuel fires of large atmospheric storage tanks - is a concern for consultees. Large scale tests for fluorine free foams are ongoing and not yet complete, partly because the scale and cost of these tests. However PFAS-free foams have provided equivalent performance to C6 foams during hydrocarbon tank fires of 15, 60 and 80m diameter (during LASFIRE testing). Performance depends on application rate and equipment, but one stakeholder suggested that there is no real reason why these results cannot be extrapolated to bigger tanks (100m) or bund fires. More testing is required to prove performance of alternatives under some conditions. To date, no realworld examples of a successful transition in installations with large tanks are identified. Consultation has noted that, as such, AFFFs are still used when large fuel areas need to be extinguished quickly or in sprinkler systems. The available evidence suggests that elsewhere technically feasible fluorine free foams have been developed, are commercially available and have been used to the satisfaction of users. This transition has not occurred without some technical challenges (and cost) and has required testing in each users' system. Additional volumes of foam, compared to PFAS-based products, have been necessary, but not uniformly. Several users have identified - and overcome - technical issues. These related to temperature tolerance of alternatives and the viscosity of foams. Some changes to foam delivery systems, nozzles and some additional storage capacity has been required. Speed of fire suppression (making allowance for additional volumes required) Limited detailed information was obtained on this specific aspect. One respondent highlighted there could be a 5-10% gap in the extinguishing time, but that this "mainly" concerned polar liquids. Other consultees noted that equivalent volumes were required and these yielded equivalent performances, but this was not consistently reported. Others noted additional volumes of fluorine free foams, compared to PFAS based products in at least some applications. Some consultees highlight that this was a particular concern with small extinguishers. Whilst one respondent noted that, in general, fluorine free foams are less flexible for users, because they have less margin for error in the proportioning (i.e. volumes required), in their application type and of ease of use. However, other consultees provided feedback of use in specific applications (aviation), including an example of where a fluorinefree foam worked satisfactorily despite deliberate inappropriate application methods as part of testing procedures. Standards The analysis of alternatives (Section E.2) provides a list of specific international compliance standards for the various commercially available products, with more details for each 181 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS shortlisted product above. The underlying study159 (Appendix 5) provides more detail on each of these standards. Foams are developed to meet specific standard requirements and it is important to note that tests used for standardisation and certification of PFAS-based foams are not necessarily appropriate for fluorine-free foams. Stakeholders highlighted during the consultation workshop that current testing protocols have often been designed with PFAS-based foams in mind. These testing protocols may not be adequately tailored to reflect the fire-fighting ability of fluorine-free foams, because the same application methods may not always be applied and read-across between different burning fuels may not be straightforward. Therefore, it is inherently challenging to compare the two types purely based on certification. Some fluorinefree foams are however capable of meeting standard firefighting certifications applicable to PFAS-based foams and this has been demonstrated in cases where some airports and municipal fire brigades for example have successfully transitioned to fluorine-free foams. 4.1.4. Use patterns of alternative fire-fighting foams to achieve comparable/acceptable performance This section discusses the impacts associated with the use patterns of alternative fire-fighting foams and includes discussion on: (a) the quantity of alternative foams needed to achieve either comparable performance or performance that is acceptable from the standpoint of safety to PFAS foams. (b) different specific application methods and equipment used. a) Quantity of foams needed to achieve comparable/best possible performance The available evidence does not permit a quantitative estimate for the comparative volumes of fluorine free foams required, for each application and with specific foams. However, the consultation allow a range to be specified. The same approach is used for the availability assessment below. It is important to note that the available quantitative information received - despite extensive attempts for specific information and for clarification - was very limited. Based on the available data, the range specified was between no change in volume and up to a maximum of 100% additional foam required, note the 100% volume estimate was specified by just one consultee and it is understood that this relates to use in one application. The available information is not sufficient to conclude these are isolated cases. As noted in the previous section, this does not apply to liquid fuel fires of large atmospheric storage tanks/large scale tank fires. Here, consultation indicates that large scale testing is still needed to confirm performance. The details on specific shortlisted products - which are known to be in use within the EU (based on stakeholder consultation) - are set out below. Table E.3 Use patterns of alternatives - summary Alternative Comparative volumes required vs PFAS containing foam Respondol ARF 3-6% No specific data has been supplied, desp te attempts to obtain this via consultation. RE-healing foam RF3X6 ATC Variable depending on applicat on ("drop in replacement, with no additional volumes required in offshore oil installations, onshore terminals and refinery). 159 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 182 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Existing use of PFAS based foams is between 14,000 and 20,000 tonnes per year. The best estimate used in the analysis is some 18,000 tonnes per year; The average price of existing PFAS-based foam is subject to uncertainty, reflecting the wide range of specific foam compounds used. It is understood that certain compounds are currently available containing high proportions of PFAS and, whilst these are judged to be effective foams, the price for these compounds is well above average, the market assessment noted that these are uncommon. The weighted average used is 3,000 per tonne of PFAS containing foam, as set out in the market assessment. Note the lowest and highest values identified in the market assessment and stakeholder consultation were 2,000 per tonne and up to 30,000 per tonne. The latter figure has a significant effect on the ranges in the socio-economic assessment, but there is insufficient data to conclude the extent of the market currently pay this price per tonne for product; Based on these parameters, the current baseline foam costs are somewhere between 28 million per year and a maximum of up to 600 million per year. The best estimate is current costs of 54million per year (i.e. 3,000 multiplied by 18,000 tonnes); The same uncertainties apply to the average prices per tonne of fluorine free foams. The market assessment concludes, based on information provided via the stakeholder consultation, that fluorine free foams, on average, are likely to be the same price, i.e. around 3,000 per tonne of foam160. This value is used in the central estimate. The ranges in the table below are the lowest and highest prices quoted in the consultation, respectively. This indicates that the most expensive fluorine free foam is likely be less expensive than the most expensive current foams. As noted above, this has a significant effect on the result and is subject to particular uncertainty; and Finally, consultees noted a range of different volumes may be required to fulfil the same/acceptable functions. The comparative volumes required differed, depending on the specific application and customer need. Therefore, a range has been used, between a 0% increase and up to 100% more fluorine free product, over and above the volumes required for PFAS-based foams. Costs for one annual cycle of foam replacement (Total EU market) Table E.4 summarises the assumptions used in the following to estimate annual foam replacement costs. Table E.4 Annual foam costs - input assumptions Baseline PFAS foam market t/yr Central (L-H) Average pr ce /tonne of foam Weighted average (L-H) Current foam costs (PFAS) EU market cost per year Best estimate (L-H) Average pr ce per tonne of foam (Fluorine free alternatives) Add t onal volumes required % increase over PFAS based foams) L-M-H 160 E.g. Eurofeu indicacted in their comment #2983 to the PFHxA restriction that the assumed pr ce level for fluorine-free foam agents of 2,200/ton might be correct for use in standard appl cations such as municipal firefighting but that the industrial grade foam agents that would be required for the chemical/petrochemical industry and similar risks are 5,000/ton. Source: Part 2 document available at https://echa.europa.eu/restrictions-under-cons deration/-/substance-rev/25419/term 184 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 18,000 (14,00020,000) 3,000 (2k-30k) 54m (28m - 600m) 3,000 (0.7k-10k) 0% - 50% - 100% Source: Market assessment, desktop research and stakeholder consultation exercise. Note, the maximum baseline used in the SEA is 21,000 tonnes, rather than the 20,000 in the market assessment due to rounding of volumes used at sector level. Using the assumptions above, Table E.5 sets out the potential costs expected to be incurred by the EU market as a whole through purchasing volumes of fluorine free foam in an annual cycle. Overall, this suggests demand for alternative foams of between 14,000 tonnes per year and up to a maximum of 40,000 tonnes per year, for the sector as a whole. The associated costs are estimated at between 21 million and 30 million per year, with c. 27 million considered to be the most likely average cost for the EU market as a whole. Again, it is recognised that individual companies/users would incur greater or lower costs per tonne and require differing volumes. The wide ranges in different foam costs indicates whilst the average company may experience some increases in costs, others would experience savings, potentially quite large savings for some very specific market segments. Table E.5 Scenarios, gross and net foam costs -annual cycle replacement costs for total EU market (M denotes millions) Costs for existing PFAS based foams (EU Market - best estimate (Range) Tonnes of alternative required (EU Market) (L-M-H) Potential foam costs using alternative products (EU Market) Best estimate (Range) Net change in foam costs (EU market) Best estimate (Range) Best estimate (assuming 18,000t) PFAS foam use p/yr) 54m (36m - 540m) 18,000 - 27,000 36,000 81m (13m to 360m) 27m (-23m to -180m) Assuming low PFAS foam use (14,000 t/yr.) 42m (28m - 480m 14,000 -21,000 - 28,000 63m (10m - 280m) 21m (-18m to -200 m) Assuming high PFAS foam use (20,000 t/yr.) 60m (40m - 600m) 20,000- 30,000 - 40,000 90m (14m- 400m) 30m (-26m to -200 m) Source: Market assessment, desktop research and stakeholder consultation exercise. It is important to note that the stakeholder consultation indicated many users had experienced no increase in foams costs and indeed no additional volumes required. The above has been undertaken to assess the potential scale in a best and worst case, using reasonable assumptions in the absence of complete data. Costs for stock write off and replacement (Total EU market) In addition to the annual replacement cycle, under RMO 2 the entire stocks of PFAS foam would need to be disposed of and alternative volumes of foam would then need to be purchased. In the baseline, foam stocks would also have to be replaced once they are used or expired, so the restriction would bring the replacement costs forward. To reflect this, the value of the depreciation of stocks at the point of replacement due to the restriction is also considered. Assuming an even age distribution of stocks of PFAS-based foam and a linear 185 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS depreciation of foams over their lifetime, the restriction would cut the life of the foams in half on average, and so half of their original purchase value would already have depreciated and should not be considered as a cost of the restriction. The restriction could also cause additional cost of purchasing alternatives instead of PFAS-based foam taking account of both the price differential and the potential difference in volumes required. As above there is significant uncertainty in the input assumptions and these are presented as a possible range of costs. Table E.6 Quantitative data - economic costs Baseline Existing stocks of PFAS-based foam: Average: 322,500 (Between 210,000 tonnes and 435,000 tonnes) Purchase costs (oneoff total for whole stock): Average: 970 million (range 420 million to 13 billion) Value of stock depreciated: Average: 485 mill on (range 210 million to 650 million) (half of purchase cost, assuming even age distribut on and linear depreciation) Restriction Scenario Volume of replacement with fluorine-free alternatives: Average: 483,750 (Between 210,000 tonnes and 870,000 tonnes) Purchase costs (oneoff total for whole stock) : Average: 1.5 billion (range 150 million to 9 bill on) Additional cost of the restriction: Average: 1.0 billion (range-60 million - 8.3 billion) (Purchase cost of replacement minus value of existing stock depreciated) Based on PFAS foam costs of 3,000 per tonne weighted average (with lowest costs of 2,000 and highest of up to 30,000 per tonne) and fluorine-free foam costs of 3,000 per tonne weighted average (with lowest costs of 700 and highest of up to 10,000 per tonne) The cost of the foam itself are only one aspect of the economic considerations of adopting alternatives. Additional transitional costs are described below. It has not been possible, despite attempts to obtain further quantitative information, to estimate costs for the market as a whole. However, several consultees noted that whilst additional costs were incurred, these were not significant and had proved manageable. Available quantitative information is summarised below. Further information is also presented in the case studies. Note that the cost of disposal of stocks of PFAS foam is covered in a later subsection (l. Costs of disposal of legacy foams). Table E.7 Quantitative data - economic costs Testing costs Storage costs Costs from technical changes Other costs including regulatory approvals No quantitative data has been obtained via stakeholder consultation, desp te several requests for such information. Testing would be associated with costs for sample volumes of foam (likely several different products) Experience in the Norwegian petrochemical sector (Equinor) included additional costs related to purchasing add t onal volumes of foam, to replace the previous PFAS containing foams, no information was provided on whether there were costs implications related Consultat on ind cated that new nozzles had been required in several cases. Typical costs for a range of firefighting nozzles are within an approximate range of between 5 or less, per piece for simple foam nozzle devices, to c. 30 and up to c.60 for marine firefighting Experience in the Norwegian petrochem cal sector (Equinor) indicates costs (labour time) in the reg on of 360,000 for a range of support in their transition at a total of 45 s tes (so in the order of c. 10,000 per site). This would therefore appear to be 186 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS for discharge, including the need for replacement nozzles; including low expansion nozzles. These challenges appear to have been caused by differences in foam viscosity. Typical costs for a range of firefighting nozzles are within an approximate range of between 5 or less per piece for simple foam nozzle devices, to c. 30 and up to c. 60 for marine firefighting nozzles or "heavy duty applicators" and up to c 200 for more specialist equipment161. Mobile foam units are in the region of 2,700.162 However, concerning large user sites like oil refineries Eurofeu and the Firefighting Foam Coalition indicated in one of their comments on the PFHxA restriction proposal163 that "the cost for handling, cleaning and retrofitting can ramp up to 15-20 Mio for a single one of Europe's biggest chemical sites. On average a site operating 20-40 storage tanks require 2-5 Mio. for retrofitting. Tankterminal.com lists 1,166 tank terminal facilities in Europe operating in total 30,982 tanks. This is per site on average 27 tanks having a total storage capacity of 215,000 m3. It also needs to be considered that any change to the system of a site also triggers additional testing and calibration costs to ensure full functionality." From Eurofeu's contribution the costs of technical changes can be inputable to elements such as the need for increased storage volume of foam concentrate and for the need to retrofit the bund areas in some tank farms to cover a significantly higher volume of liquids, associated with an emergency response action. Other costs These include regulatory approvals and those associated with bringing new products to market. Given that the market assessment noted at least some current use of fluorine free products in all sectors, further adopting fluorine-free foams would appear to be a continuity of an existing transition - so a lot of the initial costs associated with new products development will have already been incurred. Experience in the Norwegian petrochemical sector (Equinor) indicates costs (labour time) in the region of 360,000 for a range of support in their transition at a total of 45 sites (so in the order of c. 10,000 per site). Savings from adoption of fluorine free foam Many stakeholders acknowledged potential for savings from use of fluorine-free foams. The potential savings resulting from a reduction of firewater that requires disposal and hence the 161 Costs derived from search of widely available commercial products. See: https://www.made-inchina.com/products-search/hot-china-products/Water Foam Nozzle Price.html See also https://www.orbitalfasteners.co.uk/products/heavy-duty-foam-dispenser-gun See also: https://www.dortechdirect.co.uk/heavy-duty-pu-appl cator-gun-for-expandingfoam.html?gcl d=EAIaIQobChMIudKHq9GP5gIVSbDtCh31AATEEAQYASABEgL5S D BwE See: https://www.safetyshop.com/lever-operated-nozzle-for-firehose.html?gcl d=EAIaIQobChMI6cKnwNmP5gIVAuDtCh2KMwErEAQYAiABEgL91PD BwE&gclsrc=aw.ds See: https://www.fireprotectiononline.co.uk/hv-series-low-expans on-foam-branchpipes.html?gclid=EAIaIQobChMItqqg NmP5gIViLbtCh0ImgYXEAQYByABEgIFdvD BwE 162 https://simplyextinguishers.co.uk/df130-mobile-foam-units.html?gclid=EAIaIQobChMIn4yv79yP5gIVgpOzCh1PQghEAkYBiABEgKsIfD BwE 163 Comments #2983 and #3010 to the PFHxA restriction proposal, available in "Part 2" document at Part 2 document available at https://echa.europa.eu/restrictions-under-considerat on/-/substance-rev/25419/term 188 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Contamination scenarios: PFAS-containing foams and fluorine-free foams PFAS-containing foams For PFAS-containing foams at legacy sites, contamination patterns normally include soil, both unsaturated and saturated, to be impacted by PFAS at higher concentrations, because the PFAS entry point into the subsurface occurs from above ground in most scenarios, specifically for fire-fighting and training events. PFAS leaching to greater depths in the soil column by infiltrating precipitation eventually reaching groundwater is commonly observed at legacy sites. Leaching is supported by the physicochemical characteristics of PFAS. PFAS in shallow soils can also be transported via overland flow by storm water run-off during precipitation events. Storm water would either infiltrate into the ground at an area geographically separated from the original fire-fighting activities, or storm water run-off can directly discharge to a surface water body such as a river, stream, or lake, or it can be captured in a storm-/ waste water treatment facility. Historically, storm- or waste water facilities were not required to analyse for PFAS compounds. It can be assumed that most PFAS have passed untreated through a treatment works without awareness of the operator allowing for PFAS to spread to the wider environment168. The PFAS-laden soils in the source area continue to be an emission source for groundwater contamination for many years, if not decades. Once PFAS-compounds have reached the aquifer or a water-bearing unit, those compounds tend to migrate laterally and in a hydraulically downgradient direction with limited retardation from the soil matrix and negligible, if at all occurring, breakdown through biotic or abiotic processes in the aquifer169. As a consequence, PFAS tend to generate large plumes in groundwater. Acceptable PFAS threshold concentrations are extremely low, and plumes can be many kilometres long. In the Veneto region, Italy, a PFAS-production facility contaminated an area spanning more than 200 square km170. Various scenarios can result from PFAS-impacted groundwater. Groundwater could be extracted and used as drinking water. Extracted groundwater could also be used for irrigation of agricultural land. In addition to soil and groundwater impacts, surface water could be impacted from historically contaminated soils by means of surface water run-off. Under certain hydrogeological conditions, groundwater can become surface water or interact with surface water in brooks, creeks, streams, or river beds. PFAScontaminated surface water is a major concern under the Water Framework Directive with an extremely low Environmental Quality Standard (EQS) for PFOS (annual average EQS for PFOS is 0.65 ng/l)171. PFAS-impacted ground or surface water can become a challenge when they enter a water treatment works at privately owned locations (e.g. oil and gas sites or airports) or public treatment works, as indicated above. In most cases PFAS are not analysed for in water treatment works and the presence or absence of PFAS are consequentially unknown. PFAS would require in most, if not all, cases, a separate treatment step in the water treatment works with potential requirements for additional pre-treatment (e.g. high dissolved organic carbon (DOC) can be a problem in treating PFAS) and retrofitting of the treatment works at a substantial cost. 168 Nordic Council of Ministers, The Cost of Inaction - A socioeconomic analysis of environmental and health impacts linked to exposure to PFAS, 2019. 169 Concawe Report, Environmental fate and effects of poly- and perfluoroalkyl substances (PFAS), June 2016. 170 World Health Organization, Keeping our water clean: the case of water contamination in the Veneto Region, Italy, 2016. 171 Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for the Community action in the field of water policy. 193 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Fluorine-free foams Based on the definition of "remediation" and "clean-up" there would not be a remediation scenario that includes fluorine-free foam compounds as of now. For one, replacement products are fairly new to the market and possible/potential impacts from fluorine-free foams to the wider environment has not yet caused adverse effects. The current expectation is that replacement products (alternatives to PFAS-based foams) do not have the potential to contaminate soil and/or groundwater in a way that remediation can be assumed or predicted to be needed. The analysis of alternatives has shown that the substances contained in shortlisted fluorine-free alternatives (i.e. a set of alternatives considered likely to be used) exhibit lower hazards than PFAS and rapid biodegradation. Even if those alternative substances have the potential to contaminate soil and groundwater, remediation scenarios/technologies are hard to define. Remediation in most EU countries is risk-driven. That risk from alternative products cannot reasonably be anticipated at this point to develop a "remediation scenario" including treatment technologies and associated costs. There was anecdotal evidence presented by one stakeholder that fluorine-free foam caused emulsification of the run-off water in a water treatment works. Should emulsification be a recurring issue for use of fluorine-free foams, then a separate treatment step to break up the emulsion would need to be included at the water treatment works as a retrofit at an additional cost. Also, an anecdotal example was presented from another stakeholder that a permit was granted where 5000 litres of firewater runoff from fluorine-free foams could be discharged directly to a sewer after "only" a fuel separator step. Point of treatment - source area, site hydraulic control, plume, and "end-of-pipe" As described previously under fire-fighting scenarios using PFAS-containing products, PFAS compounds experience a fate and transport that can be generalised for most occurrences and described as follows (see figure172 below). At the location of the active fire-fighting activity PFAS-laden waters enter the subsurface resulting in PFAS-impacted soils - the source area (No. 1). The source area typically holds the greatest PFAS mass. Precipitation supports leaching of PFAS compounds in the unsaturated soil column to greater depth (No. 2) in the soil column eventually reaching groundwater which is then the starting point of a PFAS plume in groundwater (No. 3). Depending on the fuel that was extinguished, PFAS have a tendency to accumulate with free phase products173 at the water table intersection. The plume will extend in the direction of and grow with groundwater flow as more PFAS-mass leaches from the source area. Eventually the plume might grow to a size extending past the property boundary (airport, O&G refinery, etc.) migrating off-site. The PFAS plume size might have grown in size and extended into areas where groundwater extraction could occur for domestic (No. 9), commercial or public use (No. 10) including private drinking water wells, agricultural irrigation and livestock feeds, and drinking water production facilities. Stormwater runoff from a fire training area or live fire incident can migrate in various directions predominantly following land surface morphology (No. 5). In consequence, surface water runoff can spread PFAS contamination in directions beyond groundwater flow. Stormwater runoff can directly or indirectly occur via some sort of controlled or uncontrolled overland flow or through underground utilities. Damaged/leaking utility structures can be locations where PFAS could enter the subsurface at a point that is in only limited relation to the actual firefighting area. Stormwater or surface 172 Wood E&I Solutions, 2017. 173 Common petroleum hydrocarbon-based fuels are lighter than water (light non-aqueous phase liquids - LNAPL) and accumulate at the water table intersection when they are released to the environment at large enough quantities. "Free phase" refers to a fuel layer on the groundwater table. 194 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS with the lowest cost per kg of PFOS/PFAS removed, was calculated at 7.3 MNOK, with a cost per kg PFAS removed of NOK 357,000. Another site with the largest volume of PFOS/PFAS removed (96 kg) of all, and due to a relatively large volume of masses (22 000 m3) the total cost of remedial action was 51 MNOK, with a calculated cost per kg of the sum of PFAS removed at NOK 533,000. Clean-up is driven to a large degree by the flammable liquid itself, the soot, water and "dirt" in general terms that contribute to the fire-fighting water runoff and its potential to impact the environment. The foam used might just be another component that will need to be captured and treated, specifically under the scenario of a fluorine-free foam use. As discussed above, it is assumed that fluorine-free foams will not be persistent, mobile and toxic at levels that will require remediation (e.g. legacy site) when they enter the environment. For training activities facilities including the associated water treatment works should be engineered to account for 100% collection of all fluids including fuel and foams that the fire training water can be cleaned and treated accordingly before releasing treated waters back to the larger environment. Should a fire have been extinguished during training or a live event using PFASbased foam then it is advisable to clean-up the firefighting water promptly after the incident. Depending on the location of a live fire and the foam used, soil samples should be collected from areas where fire water runoff could have percolated into the subsurface to evaluate the presence or absence of PFAS compounds and their concentrations. Depending on the soil analytical results a need for soil exchange might be indicated. After a live fire event, regulatory communication and agreement is required for subsequent steps in the clean-up procedure to reach acceptable site conditions that will not create or leave a risk to human health or the environment. One stakeholder shared a scenario where clean-up seemed to be challenging. Fire-fighting activities in close vicinity to open water bodies (such as sea or lake) make it close to impossible to recover fire-fighting water runoff discharged into the sea or lake. To avoid runoff entering the sea, engineering solutions would be required as much as that is possible. For facilities in close proximity to large water bodies, one could possibly design berms and a drainage system to recover fire-fighting water in case a fire should truly occur. However, it might also be prudent to switch to less environmentally critical, fluorine-free, foams. Cost of remediation The study underlying this restriction proposal178 (Section 6) has assessed the typical costs of remediation of PFAS contamination resulting from the use of fire-fighting foams. The results are summarised in the table below. This shows that the typical costs per site can range from around half a million Euros (only soil remediation required, lower estimate) to just over 100 million (sum of soil excavation and incineration, groundwater pump and treat and drinking water reverse osmosis, higher estimates).179 Table E.10 Typical cost per site of remediation of PFAS contamination resulting from the use of fire-fighting foams Compartment Technique Cost 178 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 179 Please note remediation costs are highly site-specific and in certain cases can exceed the ranges provided. The estimates should therefore be considered order-of-magnitude cost ranges. 198 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Soil Excavation and off-site disposal 0.5 - 18 million Excavation and incineration 2.5 - 38 million Capping 0.42 - 4.3 million Groundwater Pump and treat 1.2 - 30.3 million Drinking water Source: Wood 2019 Reverse osmosis 2.9 - 39.8 million Section B.9. has shown that the substances contained in fluorine-free alternatives exhibit lower concern than PFAS used in fire-fighting foams, due to their lower hazards and more rapid biodegradation. On this basis, underlying study180 (Section 6) has concluded that it is currently not predicted as likely that remediation will be required as a result of the use of fluorine-free alternatives. Therefore, no remediation costs are expected to be incurred from the use of fluorine-free alternatives, implying potential savings from substitution of PFASbased foams. It is important to note that the costs refer to the remediation of legacy contamination that occurred from historical fire-fighting and/or training activities. In particular, training activities, which account for the majority of fire-fighting foam use, either already avoid the use of PFAScontaining foams and/or are conducted at contained training facilities, according to current best practice. However, the consultation did not yield information on the extent to which best practice measures are being implemented, or their effectiveness. Section B.9. has estimated that the current levels of emissions from training are likely relatively low; however historical emissions are understood to have been much higher. Fire-fighting activities typically require more immediate clean-up (discussed further in the next paragraph) rather than long-term PFAS remediation. On this basis, it seems unlikely that the current use of PFAS-containing fire-fighting foams would lead to the same remediation costs as presented for legacy contamination above. In conclusion, the restriction scenarios could eliminate the potential risk of PFAS contamination which could cause costs of up to around 100 million per site. For European sites the Nordic Council of Ministers report181 describes remediation costs associated with contamination from PFAS ranging from several hundred thousand up to 40 million with one high-cost example for the Dusseldorf Airport, Germany estimating a total remediation cost of up to 100 million. In comparison to the costs provided in this report with the Nordic report remediation costs for PFAS-impacted sites (such as airports) will total from the single digit millions to the lower double-digit millions. For Schiphol Airport 50 000 m of impacted soil were removed at a cost of 600-800/m. As described in previous sections there is a variability in costs for soil remediation depending on factors such as amount of PFAS spilled, presence of other contaminants, the volume of soil that has been contaminated, the type of soil, the environmental setting of the impacted site, and the receptors impacted or threatened. In the Nordic report cost ranges are given for three airports where costs were modelled. The modelled costs included both water and soil remediation using different methods and different levels of allowable remaining concentrations. The modelled cost ranges spanned from 2.124 million (Kristiansand Airport) over 0.4-7.1 million (Harstad/Narvik Airport) to 0.41-8.1 180 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 181 Nordic Council of Ministers, The Cost of Inaction - A socioeconomic analysis of environmental and health impacts linked to exposure to PFAS, 2019. 199 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS million (Svalbard Longyearbyen). While the remediation technologies were not reflected in the Nordic report those costs are consistent with the estimated cost range as developed in this report There are large uncertainties in the numbers of sites that may require remediation and remediation costs are very case-specific and would differ significantly across these sites, so the following estimate of total remediation costs caused by the use of PFAS-based fire-fighting foams is indicative only: The market analysis (see Annex A) estimated that there are likely to be several tens or potentially hundreds of thousands of sites that use or at least possess fire-fighting foams; If all of these would require remediation (costing some 10s of million per site), the costs of cleaning could be at most in the region of trillions of Euros (based on an assumed 100,000 sites needing remediation); However, in reality only a much smaller number of these sites would use PFAS-based foams in sufficient quantities and without adequate containment and immediate clean-up to require large scale remediation. More information on the total number of sites, real-world use of PFAS per site as well as implementation and effectiveness of best practices in terms of containment and immediate clean-up would be required to assess to which extent remediation is likely to be required in the future as a result of current use of PFAS-based fire-fighting foams; and Therefore, realistically avoided remediation costs are more likely in the order of magnitude of hundreds of millions of Euros (assuming tens of sites requiring remediation at tens of millions of Euros per site) to billions of Euros (assuming hundreds of sites requiring remediation at tens of millions of Euros per site). Cost of clean-up In addition to remediation which is driven by long-term accumulated contamination from historical releases, releases to the environment in the short-term require "clean-up". According to the stakeholder consultation, there is local or national-level regulation governing the containment or prevention of release of fire-fighting foam or firewater runoff to the environment182. One exception that has been identified is fire-fighting activities in close vicinity to open water bodies (sea, lake), where it is very difficult to recover fire-fighting water runoff discharged into the sea or lake. In the case of the lake, this could lead to remediation being required. This would relate to very specific sites in specific locations, so it would not be appropriate to estimate `typical' remediation or clean-up costs. In the case of the sea (particularly relevant for marine and offshore applications), remediation or clean-up would likely not be feasible, which raises particular concerns over the environmental impact of using PFAS-based fire-fighting foams in these applications. In all other applications, it is assumed that in most cases, the majority of fire-water run-off is contained and sent for treatment. Treatment costs for run-off can vary depending on the fire-fighting foam used: Several stakeholders that have transitioned to fluorine-free foam reported that when fluorine-free foam was used, run-off was sent to water treatment, either though the normal sewer system to the municipal WWTPs; directly to on-site waste water treatment; to other biological/chemical/mechanical 182 This was confirmed by stakeholders at least for England/Wales (The Environmental Perm tting (England and Wales) Regulations 2010 (EPR 2010)), Sweden (local author ty requirements for appl cations for new operation licenses), France (no details provided), Netherlands (no details provided), Germany ("Lschwasser-RckhalteRichtlinie" and the more detailed Bavarian "Guideline foam" which is legally binding in Bavaria and but also applied elsewhere). 200 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS treatment plants; or even drained directly to sea. One stakeholder reported that all PFAS-containing run off must be treated as a regulated waste which they do using high-temperature incineration; Stakeholders did not provide information on the cost of waste water treatment. These can vary significantly, depending on the contamination of the run-off from the flammable liquid itself, the soot and other contaminants from the fire site. For instance, UNECE 2017183 reports a cost of 1 million for disposal of 2,000 m of firewater contaminated with chemicals in a sewage treatment plant and several chemicals waste disposal facilities, resulting from a fire in a factory in Germany in 2005. This is equivalent to 0.5 per litre, or 0.64 per litre in 2019 prices184. Typical costs for regular municipal waste water treatment are much lower, for instance reported in the range of 0.0002 to 0.0005 per litre by Pajares et al. 2019185 for various municipalities in Southern Europe. Hence, treatment costs for run-off for fluorine-free foam are likely between 0.0002 per litre and around 0.64 per litre. 0.3 per litre is assumed as an average for the purpose of the approximate estimation below; Assuming that PFAS-containing run-off has to be incinerated, and assuming similar incineration costs as reported for the disposal of fire-fighting foams (see Section "l. Costs of disposal"), the costs for treatment of PFAS-containing fire-water run-off could be around 1 per litre (range 0.3 to 11 per litre). Hence, treatment costs for run-off of fluorine-free foams could be around 0.7 per litre (range ca 0-11) lower compared to PFAS-based foams186; and Data on the total amount of fire-water run-off containing fire-fighting foam per year in the EU was not available, but for illustration an example of costs per incident can be calculated. UNECE 2017187 reports five major fire-incidents in which volumes of fire-water used ranged between 2,200 and 38,000 m3. For incidents of this size, the difference in run-off treatment cost would be around 1.5-27 million (range 0-418 million) per incident188. 183 https://www.unece.org/fileadmin/DAM/env/documents/2017/TEIA/JEG MTGS/UNECE Safety Guidelines and Goo d Pract ces for Fire-water Retent on 14 Nov 2017 clean.pdf 184 2005 value converted to 2019 prices using Eurostat: HICP (2015 = 100) - annual data (average index and rate of change) (prc_hicp_aind). 185 Moral Pajares, E., Gallego Valero, L., & Romn Snchez, I. M. (2019). Cost of urban wastewater treatment and ecotaxes: Evidence from municipalities in southern Europe. Water, 11(3), 423. 186 Calculated as: Central estimate: 1/l cost of incineration of PFAS-based foams minus 0.3/l cost of waste water treatment for fluorine-free alternatives = 0.7/l cost saving; Low estimate: Waste water treatment could in some cases be more expensive (up to 0.64/l) than incineration (from 0.3/l). In these cases it is assumed that the less expensive option would be chosen and there would not be a saving of using fluorine-free foams compared to PFAS-based foams; and High estimate: The maximum possible difference is in case of the upper end of the range of incineration costs for PFAS-based foams (11/l) minus the lower end of the range of waste water treatment costs for fluorine-free alternatives (0.0002/l) 11/l cost saving. 187 https://www.unece.org/fileadmin/DAM/env/documents/2017/TEIA/JEG MTGS/UNECE Safety Gu delines and Goo d Pract ces for Fire-water Retent on 14 Nov 2017 clean.pdf 188 Calculated as: 2,200 m3 volume of fire-water run-off * 0.7/l treatment cost difference= 1.54 million.38,000 m3 volume of fire-water run-off * 0.7/l treatment cost difference= 26.6 million. These figures are rounded to two significant figures. For the wider range, instead of 0.7l treatment cost difference, 0/l (lower) and 11/l (higher) have been applied. 201 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS In cases where fire-water run-off is not contained and further clean-up is possible (i.e. runoff was not discharged to sea), there may be savings from using fluorine-free foams in terms of reduced clean-up costs: When PFAS-based foam is used and contamination of the soil and water occurs then extremely persistent chemicals are involved, which is not necessarily the case with fluorine-free foams. Stakeholders suggested in the consultation that clean-up and complex treatment is not always necessary after the use of fluorine-free foams. This could lead to potential cost savings in some cases; However, the underlying study189 (Section 6) clean-up is driven to a large degree by the flammable liquid itself, the soot, water and "dirt" in general terms that contribute to the fire-fighting water runoff, rather than the firefighting foams. Therefore, a significant difference in clean-up costs between the different types of foam used is difficult to estimate, because the incremental costs of addressing PFAS contamination is difficult to separate from the wider clean-up costs; and Clean-up costs are generally expected to be lower than remediation costs. Based on the estimates of remediation cost per site presented above, as a worst case scenario, clean-up costs can be expected to be a few hundred thousand to a few million Euros per incident. In the absence of more specific data, for illustration of the potential order of magnitude of savings: Assuming several tens of incidents per year using PFAS-based foams where clean-up is required and could be avoided if fluorine-free foams were used, the savings would be in the order of several millions to several tens of millions of Euros. 4.1.8. Availability of alternatives. This section discusses the supply-demand balance associated with a restriction on PFAS firefighting foams under RMO 1 and 2. Both scenarios will require a transition to alternatives - the difference is the speed at which this will be necessary. RMO 1 will result in a slower increase in demand as stocks are used in training and or incidents (or reach the end of their useful life) and are then replaced with new alternatives. RMO 2 will result a more sudden increase in demand as the whole market disposes of and replaces their existing stocks - potentially over a short timescale - and then require replacement stock, each year. In addition, and over and above the replacement demand, it can be assumed both scenarios will result in an increased short-term demand for testing; again the increase in demand would be greater in RMO 2 given the accelerated transition. The economic and logistical challenges of managing the transition - avoiding contamination in storage tanks and the requirements for disposal, for example - are discussed elsewhere in the SEA. Information on the specific shortlisted substances in the analysis of alternatives is summarised below - quantitative information is limited. These substances are however, illustrative and a subset of a larger range of alternative foams that are commercially available and currently in use. 189 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 202 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Table E.11 Availability of alternatives - summary Alternative Produced in the EU Currently commercially available Informat on on production volumes Respondol ARF 3- Unknown Yes 6% Not available. Stakeholders have indicated that they would not have a problem meeting increased demand in general terms. RE-healing foam Yes Yes RF3X6 ATC As above. RE-healing foam Yes Yes RF1-1% As above. Moussol FF 3X6 (F- Yes Yes 15) As above. FoamMousse 3% F- Yes Yes 15 As above. Epocol Premium Yes Yes 700 tonnes (product on and import), 500 sold in EU. Orchidex BlueFoam Yes Yes (3x3) Stakeholder (not manufacturer estimates at c.800 t/yr) Source: Market assessment, desktop research and stakeholder consultation exercise. Stakeholder consultation has provided limited information on production and use volumes of specific foams but the market assessment indicated current supply is in the region of 7,000 to 9,000 tonnes. Anecdotal information from stakeholder consultation notes that "adequate" supply exists and no consultees noted that they had experienced supply constraints in any application. Further discussions with three suppliers indicated current excess production capacity alongside additional capacity for emergencies (not quantified). The consultees noted no constraints with raw material supply. Production and sales data on one shortlisted product, Epocol, was provided as noted above in Table E.10. This data indicated total production and import capacity of 700 tonnes, with sales of 500 tonnes. Quantitative information was provided on a small number of other specific products. These are not listed above but were stated by consultees as appropriate for use in several applications, including municipal firefighting, storage facilities and marine applications. For these, total volumes produced and imported into the EU totalled a further 550 tonnes, with sales of 380 tonnes. Qualitative information on the availability alternatives was provided via stakeholder consultation on a wider range of products. A total of 22 were stated as being produced in the EU and all of these were commercially available (either in the EU, globally or both). Note that the substance identification and market assessment identified a larger number of products - in the order of 160 - but more detailed information on only a subset of these was obtained via the consultation and the assessment has focused on products for which stakeholders have indicated actual use is taking place. Using data from the market assessment, Table E.11 provides a quantitative summary of available information. First, the table provides a summary of existing EU demand for PFAS based firefighting foams. This has been split by application, based on Eurofeu survey information. Overall, this indicates current PFAS based foam demand in in the region of 18,000 tonnes per year190, with the largest use in the chemical and petrochemical sector. The 190 Note that the sum of the sectors is not equal to the total due to rounding. 203 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS second, central, column provides an overview of the volumes of alternative foams that may be expected after a restriction is imposed. This takes into account that additional volumes may be required in some applications. As in the economic feasibility section above, the analysis has been undertaken assuming no change in the volumes required (central estimate), and a 50% and up to 100% increase, respectively in the volume of foam required in all applications. It is not considered likely that this increase will be required uniformly across all applications; indeed, the stakeholder consultation indicated that many users experienced no overall increase in the volumes required. Finally, the existing demand - again based on Eurofeu survey data - Is presented on current fluorine-free foam supply in the EU. The disaggregation of demand by sector is based on the proportions specified in the Eurofeu survey. For both PFAS-based and fluorinefree foams, sector specific volumes are subject to greater uncertainty than the overall totals. Table E.12 "Top down" assessment - annual demand and supply of PFAS and Fluorine free FFF Sector of use Current PFAS foam volumes (t/yr) Central (L-H range (000's)) Existing F- free volumes (t/yr) Expected future additional demand for F-Free foams Central (L-H range 000's) Chem cal/Petrochem cal 11,000 (8-12) 2,000 - 2,600 11,000 (8-24) Municipal Fire Brigades 2,000 (2-3) 3,100-4,000 2,000 (2-6) Marine Applications 2,000 (2-2) 1,100-1,400 2,000 (2-4) Airports 2,000 (1-2) 500-600 2,000 (1-4) Mil tary 2,000 (1-2) 100-200 2,000 (1-4) Ready for use products <500 c.100 <500 Total 18,000 7,000 - 9,000 (14 - 20) Source: Market assessment, desktop research and stakeholder consultation exercise. 18,000 (14 - 40) The above information indicates that, for all uses, the volumes of fluorine free alternatives would need to increase to meet the replacement demand as users switch from PFAS containing foams under a restriction. Overall, the increase is likely to be in the order of 18,000 tonnes (i.e. sales of 18,000 tonnes of PFAS foam ceases, to be replaced by 18,000 of fluorine free foams), but potentially up to 40,000 tonnes, per year. Stakeholders indicated that spare foam production capacity exists and that users had not experienced a shortfall in supply. However, RMO 2 may result in a more sudden and potentially significantly larger demand for fluorine free foams, as existing stocks would need to be disposed of and replaced. As noted above, this could be in the region of between 210,000 tonnes and up to a theoretical maximum of 870,000 tonnes of foam. This heightens the risk of a shortfall in supply, - depending on the timescales of any restriction. Overall, the available evidence clearly indicates a range of alternative foams are currently available on a commercial basis. Moreover, data obtained from stakeholder consultation suggests that in purely quantitative terms existing production capacities can accommodate some increase in demand. For RMO 1, it has not been possible, despite further consultation attempts, to obtain quantitative information on the supply of specific products used in all 204 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS emissions on health and the environment are discussed. Emissions considered relate to both the potential for remaining PFAS compounds as well as the by-products created from disposal. The analysis focusses on the disposal of unused PFAS-containing foams, rather than the disposal of used PFAS-containing foams. Little data and information was obtained from stakeholder consultation, therefore much of this section is based on desktop research. High-temperature incineration would appear the most likely disposal option for PFAS-containing legacy foams191; Existing incineration disposal methods used apply a range of temperatures from around 400-6000C192. The literature also indicates that CF4 requires temperatures above 1,400C to decompose and that CF4 is the most difficult fluorinated organic compound to decompose193; The effectiveness of PFAS compounds to be destroyed by incineration and "the tendency for formation of fluorinated or mixed halogenated organic by-products is not well understood"194; The incomplete destruction of PFAS compounds may result in smaller PFAS or products of incomplete combustion being formed195. These products may not yet have been researched and therefore have the potential to be chemicals of concern196; The complete combustion of PFOS/PFAS will result in CO2, H2O and HF197 and the incineration of PFAS at temperatures of at least 1,100C, usually degrade PFAS to carbon dioxide and hydrogen fluoride198. It has not yet been determined what is produced when PFAS is incinerated at temperatures lower than 1,100C199; Emissions (greenhouse gases and air pollutants) from creating high temperatures for incineration: There are emissions associated with the procurement and delivery of fuel and with incinerator operation (e.g. greenhouse gases and air pollutants such as particulate matter from the combustion of fuels). Associated emissions have not been analysed and it is assumed that the incinerators would continue to operate at the same temperatures regardless of the type of waste they process. Such emissions were not highlighted by stakeholders in the consultation; Leakage during storage and transportation: Incineration processes are typically provided off-site and foams will need to be stored and transported to incineration facilities for disposal or waste equipment to be installed on-site200. During the storage and transportation of PFAS-containing foam it may be 191 Derived from stakeholder consultation responses concerning PFAS disposal methods. Note that WWT was also reported as a disposal method, but a judgement was made that these disposal techniques relate to used PFAS-containing firefighting foam rather than unused foam. 192 Obtained from stakeholder consultation. 193 https://www.epa.gov/s tes/product on/files/201909/documents/technical brief pfas incineration oaa approved final july 2019.pdf 194 https://www.epa.gov/s tes/product on/files/201909/documents/technical brief pfas incineration oaa approved final july 2019.pdf 195 https://www.epa.gov/s tes/product on/files/201909/documents/technical brief pfas incineration oaa approved final july 2019.pdf 196 https://www.epa.gov/s tes/product on/files/201909/documents/technical brief pfas incineration oaa approved final july 2019.pdf 197 https://www.diva-portal.org/smash/get/diva2:1155115/FULLTEXT01.pdf 198 UNEP, 2012 in: https://www.kemi.se/global/rapporter/2016/report-11-16-strategy-for-reducing-the-use-of-higlyfluorinated-substances-pfas.pdf 199 https://www.kemi.se/global/rapporter/2016/report-11-16-strategy-for-reducing-the-use-of-higly-fluorinatedsubstances-pfas.pdf 200 https://www.serdp-estcp.org/content/download/48955/466822/file/ER18-1593%20Final%20Report.pdf 206 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.2.1. Physical destruction - Incineration in Hazardous Waste Incineration plants 4.2.1.1.1. Background The strongest bond in a PFAS-molecule is the carbon-fluorine bond with a bond strength of 485 kJ/mol (Roesch et al. 2020). This bond needs to be broken in order to completely destroy a PFAS molecule. The breaking of only the carbon-carbon bonds may lead to the formation of shorter fluorinated molecules, such as ultra-short chain PFAS like trifluoroacetic acid and fluorinated gases like hexafluoro ethane (C2F6) and tetrafluoro methane (CF4). The complete thermal destruction, meaning mineralisation, of a PFAS molecule leads to hydrogen fluoride, water, and CO2. 4.2.1.1.2. Technical performance The main principle of waste incineration lies in the thermal breaking of the chemical bonds in a molecule. For this the IED requires European waste incinerators to operate at a minimum temperature of 850C with a residence time of at least two seconds. In Europe, for hazardous waste with more than 1 % of halogenated organic substances (what would also apply to PFAS-based firefighting foam run-off and cleaning water) the incinerator needs to reach temperatures of at least 1,100C (2010/75/EU 2010). The respective incinerators are commonly called hazardous waste incinerators (HWI). To current knowledge, the conditions can break the chemical bonds of a molecule and transform the waste into CO2, water, salt, and ash. Hazardous waste incinerators are designed to handle and destroy the most difficult hazardous (explosive and/or toxic) substances. Hazardous waste incinerators have specialized systems for the input of waste material, depending on the type of waste being handled. This is particularly important for some of the most hazardous and toxic wastes. Options include a solid waste bunker, a tank farm for liquid and pasty wastes, drum storage and transportation facilities. For certain (highly reactive) wastes, a dedicated direct injection system is necessary. The decomposition temperatures for PFAS vary depending on chain length and functional group. PFOA decomposes already at around 100 C, FOSA at 150 C, PFHxS and PFOS around 350 C and PTFE at around 500 C. At these temperatures the bonds inside the compounds are broken and gaseous fragments are formed. During the decomposition of PTFE fragments such as CF, CF3, C2F4 and C3F5 can be found which indicates, that not all carbon-fluorine bonds were broken (Wang et al. 2015). To completely mineralise PFAS to hydrogen fluoride, water and CO2 higher temperatures are needed. According to current literature the temperatures should reach at least 1,100 C to degrade PFAS to carbon dioxide and hydrogen fluoride (KEMI 2016). The Danish Ministry of Environment published a report on the incineration of persistent organic pollutants (POPs) including PFOS. It is stated, that PFOS will be destroyed to more than 99 % by co-incineration and that other studies have shown a destruction efficiency of more than 99.97 % for fluorotelomers, chlorofluorocarbons and PTFE in conventional waste incineration. It is however also stated, that during the decomposition of PFOS at 900 C simple fluorocarbons such as CF4, C2F6, CHF3 and C2H2F2 will be formed (Lundin & Jansson 2017). Among the fluorinated gases tetrafluoro methane (CF4) is the hardest to destroy, as it only contains carbon-fluorine bonds. 208 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 3: The hazardous waste incinerator process taken from Eurits homepage 4.2.1.1.3. Side products and emissions A study performed in 2014 analysed the ash and waters arising from multiple waste incinerators in Sweden. The incinerators generally operated at temperatures above 850 C and employed a flue gas cleaning process where the flue gas is first pumped through an acidic solution and then through a neutral step where sulphur dioxides are separated through the addition of lye. While multiple PFAS could be found in all sampled media the authors conclude that as the amounts were so low that waste incineration plants in Sweden are unlikely to contribute significantly to environmental emissions of PFAA (Sandblom 2014). According to data from the US, the end product of the complete combustion of any organic compound will lead to carbon dioxide and water which will be emitted to the air. In the case of PFAS, hydrogen fluoride will also be formed if the compound is completely destroyed. It can be found in the bottom ash as well as the flue gas. In order to remove the HF from the gas a gas scrubber is applied. For this the hot flue gas is cooled in a quenching unit filled with water whereby the HF dissolves in the water. This step is then repeated with a multistep scrubbing tower where the flue gas is scrubbed with a sodium hydroxide solution to remove all remaining HF. The resulting effluent is then quenched in a calcium hydroxide solution where the dissolved fluorine precipitates as calcium fluoride. As this method employs an alkaline solution it may also remove any airborne charged PFAS such as PFCA and PFSA but can however not remove fluorinated gases(US-EPA 2020a). If the temperature is too low products of incomplete destruction will be formed. These include for example CF4, C2F6, CHF3, C2H2F2 and C3F8(DK-EPA 2019; US-EPA 2020a). A quantitative analysis of the formation of these fluorinated gases during the combustion of PFAS has not yet been performed. In general, these gases have a high greenhouse gas potential and should be avoided. As the PFAS destruction efficiency of the thermal treatment is not 100 %, small amounts of PFAS will not be destroyed and as such can be emitted to air or be found in the fly and bottom ash. The ashes are often landfilled and the contained PFAS 209 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS can thus be washed out and emitted into the water and ground. However, based on data from Sweden, PFAS concentrations in fly and bottom ash are very low (26 - 748.3 pg/g) (Wohlin 2020). Another study concluded that in total less than 10 kg of PFAA are deposited on Swedish landfills per year from ash from waste incinerators (Sandblom 2014). Data from stakeholder interviews indicate that there is a need for standardisation and future scientific investigations: One stakeholder from Germany made it clear that there is still a need for research with regard to the incineration of PFAS-containing wastes and the associated issues, particularly with regard to the required minimum temperatures and possible products of incomplete incineration. In the past, investigations have already been carried out, for example at household waste incineration plants, but these often focused on individual substances such as PFOA and PFOS or long-chain compounds. Although it can be assumed that these compounds break down at sufficiently high temperatures and long residence times, the extent to which short-chain compounds or products of incomplete combustion (PICs) are formed or emitted and how these are to be evaluated has not yet been sufficiently researched according to current knowledge (LASTFIRE-Interview 2021). Another stakeholder from Germany stated that currently measured background levels of PFAS substances must come via incineration. According to measurements in Bavaria, when PFAS is measured in soil 50% of taken samples would be over current threshold levels as defined by the Bavarian PFC assessment guidelines (measurements based on DIN 38414-14) (LfU-Gierig-Interview 2021). The same stakeholder also indicated that so far, there are no validated measurement methods for the determination of PFAS in exhaust air. However, a DIN-standard for the determination of PFAS in exhaust air is drafted right now. The stakeholder guessed it will take approx. 2-3 years to publish it (LfU-Gierig-Interview 2021). 4.2.1.1.4. Availability across the EU According to the Nordics Council of Ministers there are 808 incineration facilities in Europe, including hazardous and municipal waste incinerators (NordicCouncil 2019). The Confederation of European Waste-of-Energy Plants reported in 2018 that there are 492 waste to energy plants operating in Europe. This number does not include the hazardous waste incineration plants. In total the 492 plants treated 96 million tonnes of waste in 2018 (CEWEP 2018)204. Hazardous Waste Europe represents 155 hazardous waste treatment installations in Europe with a total treatment capacity of 4.6 million tonnes per year. These facilities however also include non-incineration processes such as biological treatment and landfills205. Another association, the European Union for Responsible Treatment of Special Waste - EURITS, shows on its homepage the availability of HWI across the EU (see Figure 2). Of course, this overview only includes member companies of this respective association. Based on the overview it can be assumed, that HWI availability differs across Europe. This is in line with a stakeholder comment from Norway who reported that there is no HWI available 204 Assuming that the difference between the 808 incineration facilities and the 492 waste to energy plants are hazardous waste incinerators it could be assumed, that there are 316 hazardous waste incinerators in Europe. However, the exact numbers are not known. 205 Numbers are taken from the respective homepage, see here accessed at 01.04.2021 210 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS in Norway, thus PFAS-based AFFF was sent to cement kilns (see also next chapter) (EquinorYstanes-Interview 2021). Another stakeholder from the Netherlands indicated that there would be no such incineration plant in the Netherlands. Neighbouring countries Belgium and Germany would have these (LEC-BrandweerBRZO-Submission 2021). Figure 4: Availability of HWI across the EU taken from Eurits homepage According to the German Federal Environmental Agency there are 29 hazardous waste incinerators in Germany with a total capacity of 1,520,490 million tonnes per year206. The WI BREF reported 121 hazardous waste incinerators in Europe in 2019 with a total capacity of 6.75 million tonnes of waste per year however the exact incineration conditions are unknown. Costs207 According to the Wood's report the cost to incinerate one PFAS-containing AFFF litre range between 0.3 - 1.5 /l (Wood et al. 2020). This range is in accordance to the data gathered in the stakeholder engagement and literature review of this project: On their website, the Rosenbauer Group reports a price of 200 - 400 /m for the high temperature disposal of PFAS-containing AFFF, which corresponds to 0.2 - 0.4 /l208. 206 Numbers are taken from the respective homepage, see here accessed at 01.04.2021 207 The following assumptions have been considered: density of PFAS-containing AFFF is approximated to be 1 kg/m3 and exchange rate Euro to US dollar of around 1,2:1 (as of 01st of April 2021) 208 Numbers are taken from the respective homepage, see here accessed at 01.04.2021 211 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Also, another company from Germany offers to take back foams for 1 - 2 /l209. A stakeholder from Germany named a price of 700 /t for the incineration of PFAScontaining AFFF which corresponds to 0.7 /l (DUS-Valentin-Interview 2021). Another stakeholder from Germany named a price of about 400 - 600 /t, which corresponds to 0.4-0.6 /l (LfU-Gierig-Interview 2021). Another stakeholder from Germany named prices between 700-1000 /t, which corresponds to 0.7-1 /l. The specification of the fluorine content before incineration is obligatory Interview 2021). Higher prices are reported for the US and Australia. The US EPA published an interim guide on the destruction of PFAS where a price for liquid halogenated hazardous waste of 1,218 - 1,770 $/ton is stated. This corresponds to roughly 1 - 1.46 /l (US-EPA 2020b). For Australia 2,000 per m are reported, which corresponds to 2 /l210. 4.2.1.1.5. Additional information and available case studies According to one stakeholder from Germany, incineration plans often do not accept PFASbased AFFF, because of its foaming capacities (the liquid waste is fed into the combustion chamber through a nozzle) and the formation of HF-acid (corrodes the tiling). This could lead to the fact that the prices for AFFF-incinerations will increase in the future ( Interview 2021). On stakeholder from Germany stated that the only publicly accessible plant for a thermal treatment of waste containing PFCs in Bavaria is GSB - Sonderabfall-Entsorgung Bayern GmbH in Ebenhausen near Ingolstadt. The incineration plant consists of 2 lines with a total annual throughput of approximately 220,000 tons. In 2020, GSB thermally disposed of about 834 t of waste from the segment of extinguishing foam, extinguishing water, extinguishing agents, for which at least a PFC contamination could not be excluded in principle; only 23.88 t of foam extinguishing agents, extinguishing foam, extinguishing water contain a specific reference to PFC or PFT in the waste designation. Since the exhaust gas cleaning technology used in Ebenhausen consists, among other things, of various scrubber stages, which generally ensured a high separation of halogenated pollutant compounds such as HF, the emission of HF is far below the legal limit. Combustion temperatures average 1080C in the rotary kiln and 1000C in the afterburner chamber. Thermal destruction of components containing PFC/PFT can therefore be assumed with a high degree of probability (LfU-Gierig-Interview 2021). A stakeholder from the Netherlands brought up the idea to set up consortia in different regions in Europe for the destruction of foam concentrates. Many parties (public and private) will soon have foam concentrates that may no longer be used. Tackling this together seems a good option for cost-efficient and environmentally friendly solutions (LEC-BrandweerBRZO-Submission 2021). 4.2.2. Physical destruction - Incineration in cement kilns 4.2.2.1.1. Background According to (Lundin & Jansson 2017), cement kilns typically consist of a long cylinder of 50- 150 meters in length, inclined slightly from the horizontal (3% to 4% gradient), which is rotated at about 1-4 revolutions per minute. Raw materials such as limestone, silica, alumina, 209 Numbers are taken from the respective homepage, see here accessed at 01.04.2021 210 See comment on Rosenbauer homepage, see here accessed at 01.04.2021 212 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS and iron oxides are fed into the upper or "cold" end of the rotary kiln. The slope and rotation cause the materials to move toward the lower or "hot" end of the kiln. The kiln is fired at the lower end, where material temperatures reach 1,400C-1,500C. The fuel used to heat the rotary kiln has traditionally been coal, but lately different kinds of waste fractions have been utilized in some plants. Wang et al. published a paper in 2015 indicating, that the addition of calcium hydroxide can catalyse the defluorination process of PFAS. At temperatures of 900 C this method showed high transformation rates, indicated by the formation of calcium fluoride. For PFOS a transformation rate of 90 % was achieved with even better results for PFHxS. PFOA and FOSA however only reached transformation ratios of around 50 % suggesting, that the functional group has an influence on the efficacy of the method. PTFE reached transformation ratios of 80 % already at a temperature of 400 C (Wang et al. 2015). Comparing this to the decomposition temperature of 500 C for PTFE, the calcium salts can lower the needed reaction temperature by 100 C. This research suggests that the addition of these salts to the incineration process can lower the formation of fluorinated gases. According to data from Australia, the advantage of adding PFAS waste to the production of clinker in cement kilns, is that no extra energy is required to destroy the PFAS and additionally the quality of the clinker can be enhanced through the addition of fluorine (Holmes & Queensland 2020b). Fluorinated substances react in the cement production process as mineralizers, which can promote the formation of a specific phase altering the thermodynamic equilibrium of reactions. Mineralizers are more efficient in the presence of a liquid phase and can contribute to the flux activity. Next to fluor other examples are: Zinc, Manganese, Sulphur, among others (Cemex 2013). The addition of fluoride has proven to increase the reactivity of clinker used in cement as well as reducing the amount of clinker needed. Typical fluorine addition rates are 0.2 % by weight of clinker to achieve mineralisation without adverse effects (Cooper 2014). Fluorine is often added in the form of calcium fluoride to the cement kiln but can also be added in the form of PFAS, however the calcium fluoride content should not be lower than 40%. The inclusion of calcium fluoride can decrease the burning temperature by 100 C (Cemex 2013). As limestone (calcium and magnesium carbonate) is an ingredient for the production of clinker PFAS could be added to form in situ calcium fluoride (CaF2) in the cement kiln. The preferred method of introduction of PFAS wastes is by blending the foam concentrate or any other liquid wastes into the alternative fuels (waste oils) so as to control and minimise the water content that would otherwise disturb the temperature of the burner flame. Solid wastes such as PFAS contaminated GAC and resins can also be introduced packaged in 20L buckets at a suitable point in the kiln as is currently done for clinical and drug wastes. Overall, it is considered that the use of cement kilns for PFAS destruction represents the best option based on the very large safety margins in the normal production conditions for complete destruction (calcium, high temperature, long residence times), permanent capture of the fluorine as inert, insoluble, non-toxic minerals, no need to modify kiln equipment, and no need for additional fuel/energy costs (Holmes & Queensland 2020a). 4.2.2.1.2. Technical performance The cement kiln generally operates at temperatures between 800 - 1,800 C depending on which process step with a total residence time of about 25 minutes. At the hottest point the residence time at ~17 - 21 seconds at 1,800 C (Holmes & Queensland 2020b), which according to recent literature is hot enough to even destroy CF4. As such this technology can be used to effectively destroy PFAS and at the same time produce cement clinker. The Queensland government in Australia has already conducted a trial run with a total fluorine input of 325kg/h from which 5kg/h was from PFAS. As a result, no PFAS and only minimal amounts of hydrogen fluoride could be detected after the burning process. The quality of the clinker was unaffected (Holmes & Queensland 2020b). 213 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Also, according to US-EPA, the temperature at which the cement kilns operate (usually around 1400C-2000 C) allows for full destruction of PFAS compounds and the residence time (610 sec) is believed to be sufficient (Patterson & Dastgheib 2020). For conventional waste incinerators on average ~354 - 534 kWh/m of energy is needed at 1,100 C for the burning of waste (Holmes & Queensland 2020b). Maga et al. state an energy demand for the high temperature incineration of AFFF containing spent fire-extinguishing water at 1,100 C of 1,312 kWh/m (Maga et al. 2021a). This value is higher due to the added energy needed to vaporise the water. 4.2.2.1.3. Side products and emissions According to stakeholder knowledge, cement kilns do not possess the same filter techniques as incinerators handling hazardous waste (HWI). This needs to be considered when emissions are discussed (DUS-Valentin-Interview 2021). However, there are no standardized methods to monitor PFAS in exhaust air from incinerators as discussed above. Other Stakeholders also indicated, that there is no knowledge about a possible PFAScontamination (or other fluorinated side products) of the end product (cement) (LASTFIREInterview 2021). Data from Australia however, indicate that when PFAS introduced to both the main burner and the calciner produced results of very high destruction efficiencies with no PFAS in flue gases and no change to the usual emissions of very low levels of HF in normal clinker production. A significant point to note is that the trial runs of destruction of PFAS at 5kg/hr (as F) were done with and without the input of aluminium smelter cell waste materials with fluorine throughput of 325kg/hr (as F). The destruction of the fluorine-containing (~15%) spent cell carbon and refractory waste has been common practice in cement kilns for decades with low HF emissions demonstrating the very high efficiency of fluorine capture by calcium and the failure of the carbon and fluorine to recombine into PFAS. The Cement Australia kiln at Gladstone is licensed to destroy up to 5kg/hr PFAS (as F) based on the maximum throughput rate in the trials (at ~4.50/L). However, the other larger fluorine inputs with no significant HF or PFAS outputs suggest that greater throughputs of PFAS wastes could be destroyed just as effectively as the 5kg/hr (as F) in the licence (Holmes & Queensland 2020a).. 4.2.2.1.4. Availability across the EU According to the best available techniques reference document for the production of cement, lime and magnesium oxide, there are 268 cement kilns in Europe. In 2004 6.1 million tonnes of waste was used as fuel in cement kilns from which one million tonnes were hazardous waste. It is also stated that in 2007 17% of fuels was sourced from waste (CLM-BREF 2013). German authorities are not aware that the incineration of PFAS-based foams in cement kilns are taking place in Germany (DUS-Valentin-Interview 2021; LASTFIRE-Interview 2021; LfUGierig-Interview 2021). According to other stakeholders this is a developing field in the EU (LASTFIRE-Interview 2021). Based on desktop search, also no other cases are reported. However, in Australia calcium catalysed destruction in cement kilns is currently best practice (Holmes 2020) . 4.2.2.1.5. Costs One stakeholder indicated costs for incineration in cement kiln in Norway of 1-2$/litre, what would correspond to 0.85 to 1.7 /l (Equinor-Ystanes-Interview 2021). Australian Stakeholder indicate a cost of 4.50/L (Holmes & Queensland 2020a). This price (EU-based) is comparable to the prices reported for HWIs (0.2 - 2 /l). 4.2.2.1.6. Additional information and available case studies 214 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS One stakeholder sent its waste to a cement kiln (Norcem in Brevik) in Norway, which uses temperatures of 2,000 C. To his knowledge this would be the only waste disposal option in Norway, as most municipal waste incinerators operate at lower temperatures (800 C) (Equinor-Ystanes-Interview 2021). 4.3. Conclusion on the disposal of PFAS-containing firefighting foam concentrates The incineration of PFAS-containing AFFF is the most used disposal method. Literature indicates that waste incinerators at temperatures of 900 C are able to destroy PFOS at more than 99%. A destruction efficiency of more than 99.97 % for fluorotelomers, chlorofluorocarbons and PTFE in conventional waste incineration was also reported. However, this process might not lead to the complete mineralisation of the PFAS i.e. the decomposition of the PFAS to CO2, water, and hydrogen fluoride. At these temperatures short chain fluorinated compounds such as CF4, C2F6, CHF3, C2H2F2 and C3F8 can be formed and released to the air. Literature indicates that temperatures of at least 1,400 C are needed to destroy CF4 and as such completely mineralise the PFAS. Literature indicates, that 1,100 C is sufficiently hot and feasible for the destruction of PFAS, however no study has provided quantitative results on possible fluorinated gas emissions. The average cost of approximately 1/l (range is 0.2-2 /l) is comparatively cheap but the process requires high amounts of energy as the water needs to be vapourised. No actual data has been found that would indicate that the cost for incineration increased recently or will increase in the future. However, based on input of a stakeholder, the capacity is an issue for some hazardous waste incinerators, as not all of them can handle large amounts of liquid waste and foaming can cause issues, when it is stored intermediately with other liquid waste (WFVD & 2021). Also, according to another stakeholder from Germany, incineration plans often do not accept PFAS-based AFFF, because of its foaming capacities (the liquid waste is fed into the combustion chamber through a nozzle) and the formation of HF-acid (corrodes the tiling). This could lead to the fact that the prices for AFFF-incinerations will increase in the future ( Interview 2021). The co-incineration of PFAS waste in cement kilns is a viable alternative to incineration in HWI, as these kilns reach temperatures of up to 1,800 C with residence times of ~20 seconds. It has been shown that the addition of calcium fluoride can increase the quality of the clinker. Additionally, calcium salts can decrease the decomposition temperature of PFAS and increase the mineralisation rate by forming calcium fluoride. Through the addition of PFAS-containing waste to the clinker production in-situ calcium fluoride can be formed, which can increase the clinker quality and destroy the PFAS. The applicability of liquid AFFF concentrate in the cement kilns in the EU is rather unclear. German authorities are not aware that the incineration of PFAS-based foams in cement kilns are taking place in Germany (DUSValentin-Interview 2021; LfU-Gierig-Interview 2021). One stakeholder from Norway indicated that his company sent PFAS-based firefighting foams to a cement kiln as there in no HWI available in Norway (Equinor-Ystanes-Interview 2021). No costs have been reported for this case. In Australia calcium catalysed destruction in cement kilns is well established and currently best practice (Holmes 2020). Australian Stakeholder indicate a cost of 4.50/L (Holmes & Queensland 2020a). However, stakeholders from Germany indicated, that cement kilns don't have the same filter techniques as HWIs (DUS-Valentin-Interview 2021). In addition, stakeholders from Germany are concerned that the cement could also be contaminated (DUS-Valentin-Interview 2021). However, data from Australia indicate no contamination of the cement and a very high destruction efficiencies with no PFAS in flue gases and no change to the usual emissions of very low levels of HF in normal clinker production (Holmes & Queensland 2020a). According to German federal environmental authorities the degree of destruction of PFASs (e.g. related to the input concentration) during incineration is not well understood. In general, there is still a need for research concerning the incineration of PFAS-containing wastes and thus also of AFFF concentrates (LANUV-Voland-Response 2021). 215 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS has transitioned to fluorine-free foams also reported that their stocks of PFAS-based foams were incinerated in a cement kiln, but reported costs around 1 per litre. Plasma arc furnace conditions can reach 6000C and have an estimated processing cost of 11/l.211 It would therefore appear that costs are higher for incinerators operating at higher temperatures and there is a potential trade-off between the cost of incineration, effectiveness of PFAS destruction and time, due to higher temperatures being more likely to completely destroy the PFAS. With the sudden increase in short-term demand for incineration, existing disposal methods would need to be sufficient to process the volume of legacy PFAS foams required to be disposed of. Where capacity is insufficient, the storage of the foam will likely be required. The following assumptions are made to derive the capacity for existing incinerators to process PFAS-containing foam and the time it would take to complete this (not taking into account transportation times): The literature indicates that there are 808 incineration facilities across EU28212. These include high temperature hazardous waste incinerators as well as municipal waste incinerators that probably operate at lower temperatures. However, according to the Industrial Emissions Directive213 Article 50, all incinerators need to be designed, equipped, built and operated so that a temperature of at least 850C is achieved for at least two seconds. It is therefore assumed that all 808 incinerators are able to operate at least at 850C. However, as discussed in the previous sub-section, at least 1,100C (or for some PFAS even at least 1,400C) are required to degrade PFAS to carbon dioxide and hydrogen fluoride and it has not yet been determined what is produced when PFAS is incinerated at lower temperatures. Data was not available to determine the share of EU incineration facilities that achieves 1,100-1,400C; The amount of PFAS-containing legacy foam for disposal is between 210,000 tonnes and 435,000 tonnes (average 322,500 tonnes); Information obtained from stakeholder consultation indicates that an incinerator operating at around 850-1800C can process one tonne of foam per hour and an incinerator operating at around 6000C has a throughput of around 25l per hour. It is assumed that 1kg = 1l; and It is also assumed that incinerators continuously operate with the same processing capacity and at the same temperature, 24 hours a day. Based on the above assumptions, the tables below provide estimates of the time it will take incinerators to dispose of fire-fighting foams based on 808 incinerators having a processing capacity of 25l per hour or one tonne per hour. As discussed above, to ensure adequate destruction of PFAS, it would appear to be preferable to dispose of PFAS-containing firefighting foams at incinerators with higher temperatures (at least 1,100-1,400C). This will therefore reduce the capacity available and increase the time period required for disposal. 211 Obtained from stakeholder consultat on. Note that t is not clear whether this relates to foam concentrate or other foam types being processed. 212 The Cost of Inaction - http://norden.diva-portal.org/smash/get/diva2:1295959/FULLTEXT01.pdf 213 Directive 2010/75/EU of the European Parliament and of the Council of 24 November 2010 on industrial emissions (integrated pollution prevention and control) 217 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Table E.13 Processing time based on existing incinerator capacity processing 25l per hour. Foam to be disposed of (tonnes) Time for foam to be disposed of (hours) Time for foam to be disposed of (days) 210,000 (low) 10,400 400 322,500 (average) 16,000 700 435,000 (high) 21,500 900 Note that all cost values are assumed to represent the cost of disposal of unused PFAS-containing fire-fighting foams and not used PFAS-containing firefighting foam. Source: market assessment, desktop research and stakeholder consultation. Values have been rounded. Table E.14 Processing time based on existing incinerator capacity processing one tonne per hour. Foam to be disposed of (tonnes) Time for foam to be disposed of (hours) Time for foam to be disposed of (days) 210,000 (low) 260 10 322,500 (average) 400 20 435,000 (high) 540 20 Note that all cost values are assumed to represent the cost of disposal of unused PFAS-containing fire-fighting foams and not used PFAS-containing firefighting foam. Source: market assessment, desktop research and stakeholder consultation. Values have been rounded. Responses from the stakeholder consultation indicate that there is sufficient capacity for disposal of PFAS-containing foams. One stakeholder reports that there is sufficient capacity for disposal, but that getting high-temperature incineration capacities is becoming more difficult. Another stakeholder also reports that sufficient capacity for disposal by incineration is not guaranteed. Capacity for disposal is also likely to depend on the transition period chosen (yet to be determined) and was mentioned in the stakeholder consultation. If the transition period is short, there is the potential for demand for disposal facilities to outstrip supply. A longer transition period is more likely to result in the demand and the quantities sent for disposal being spread over a greater time period. Alternatively, a sector by sector introduction of RMO 2 could be introduced to also spread the demand for disposal over time and avoid destruction capacity being exceeded214. Note that it is not clear whether stakeholder responses relate to used foams or whether responses relate to the sufficient capacity for the disposal of legacy foams if RMO 2 were to occur. Additionally, the geographical locations of incinerators, the feasibility of storing and transporting PFAS to destruction facilities as well as the availability of transportation vehicles and labour has not been evaluated due to lack of information from both desktop-based research and stakeholder consultation. Further, the knock -on effects on other sectors requiring use of incineration facilities have not been considered. 4.3.2. Costs of disposal of legacy foams This section discusses the costs associated with the disposal of legacy PFAS-containing firefighting foams under RMO 2. Costs occur from the disposal process itself, as well as from transportation to disposal facilities and the storage of PFAS-foams. Costs of disposal are not considered to be incurred in RMO 1, unless `transitional wastage' occurs, where the disposal of some PFAS-containing foam must happen to enable a switch to an alternative. Information 214 Obtained from stakeholder consultation. 218 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS According to JOIFF, from a waste management perspective, treating foam concentrates and spent foam mixtures resulting from AFFF and fluoroprotein foams used in fire incidents is not possible using biological treatment processes. Conventional wastewater treatment plants will not breakdown non-biodegradable PFASs. Discharge of these wastes to sewer is therefore not an effective treatment (JOIFFF 2020). In 2020, UBA together with Arcadis highlighted in a review article all available PFAS treatment technologies for groundwater and ranged them according to their practicality (UBA 2020). In Figure 3 a visual summary of this overview is shown. Figure 5: PFAS treatment technologies for water, ranged according to their practicality (taken from (UBA 2020)). When PFAS treatment technologies for water are discussed the volume of the water and the proportional PFAS-concentration need to be considered. According to Horst et al., the current state of the practice for treating water contaminated with PFAS is to take extremely large volumes with low PFAS concentrations - typically in the part per trillion range (ppt; i.e., nanogram per litre [ng/L]); and convert it into much smaller volumes of high PFAS concentration, which can then be more economically treated using technologies attempting to destroy PFAS (Horst et al. 2020). In Figure 4 the conceptual impact of volume on the relevance of currently available non-destructive and destructive treatment approaches for PFAS contaminated water is shown. 221 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 6: Conceptual impact of volume on the relevance of currently available nondestructive and destructive treatment approaches for PFAS contaminated water (taken from (Horst et al. 2020)) Fire run off water and cleaning water In the context of this project, only feasible and mature techniques for the treatment of fire run off and cleaning water (shown in Figure 5 in the green box) have been analysed in detail. For both PFAS-contaminated water types it is assumed that a rather high PFAS-concentration is to be expected. For example, PerfluorAd is designed for treating water containing PFAS concentrations greater than 0.3 g/L (Ross et al. 2018). Figure 7: PFAS treatment technologies for water, ranged according to their practicality (taken from (UBA 2020)). Available techniques focussing on in situ techniques for groundwater are not considered (e.g. activated carbon injection into aquifer218), as they lack market maturity and are not compatible with both types of water (Concawe 2020). 218218 According to the national geographic society an aquifer is an underground layer of water-bearing permeable rock, rock fractures or unconsolidated materials (gravel, sand, or silt). Groundwater can be 222 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS In addition, those technique for which no references were available for treatment of fireextinguishing waters or water with PFAS concentrations within the range of fire-extinguishing waters have been also not analysed. According to Concawe, for the following techniques there are no reported case studies regarding fire-extinguishing water treatment: Electrochemical degradation Sono-chemistry UV-radiation Plasma treatment These non-destructive and destructive treatment techniques have therefore not been considered. The techniques have been analysed and updated based on current available literature (predominantly the ITRC-guideline (ITRC 2020) and a review of water treatment systems for PFAS removal from Concawe (Concawe 2020)) and finally stakeholder input. According to UBA, the treatment of high AFFF-contaminated water poses a challenge. With the help of electrocoagulation and filtration, the water was prepared to such an extent that it could be treated by reverse osmosis (degree of purification approx. 99.9%) (UBA 2020). 4.4.1.1.2. What happens to fire-run off water? One stakeholder from Germany indicated, that PFAS-contaminated fire-run off waters mostly enter the environment (both via WWTP and directly) and (company-owned or municipal) WWTPs. Those who use chemical and physical treatment methods only are not suited to appropriately handle PFAS. In his opinion, a more suited way of handling the run-off water would be to collect it and store it in silos, where it can be treated. However, he observed this only in rare cases. Legally, in Germany, the run-off water after an incident is the responsibility of the company in which the fire occurred. Based on an article by three cases are to be distinguished when the fate of run- off water is to be characterised ( 2021): 1. If the fire event occurs on unsealed surfaces and/or grounds that do not have retention facilities or catchment areas for the extinguishing water, it must be assumed that the extinguishing water will infiltrate into the subsoil and possibly also into the groundwater (see Figure 6). Following infiltration, the contaminated soil material may have to be excavated and then disposed of (e.g. landfilled or incinerated), as shown in Figure 7 or the groundwater may have to be cleaned up over many years by means of a pump-and-treat measure (see Figure 7). 2. If the fire occurs on a paved area and the extinguishing water flows directly to the natural receiving water via the storm drain system, there is no possibility of intervention and the environmental impact is immediate. If, however, the water enters a sewage system, it might be possible to collect the PFAS-contaminated extinguishing water in the basin systems of the wastewater treatment plant. For this, the necessary space would have to be available, the "wave of pollutants" would have to be collected in a targeted manner and diverted into the buffer basins. If this is not possible - which is likely to be the more frequent case in practice - then it must be assumed that a significant share of the PFAS substances will pass through the wastewater treatment plants without any targeted treatment of these non-biodegradable substances. 3. Companies that are subject to the Extinguishing Water Retention Directive (LRRL) have bunding areas in which the extinguishing water can be temporarily stored. In the case of intermediate storage on site, various options can be selected for the subsequent handling of PFAS-contaminated extinguishing water. extracted using a water well. The study of water flow in aquifers and the characterization of aquifers is called hydrogeology. See here, accessed at 02.04.2021. 223 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS a. Transport of the extinguishing water in silo vehicles to off-site water treatment plants: The so-called CP plants (chemical-physical water treatment plants) are mostly plants that accept process waters from trade and industry. In many cases, pre-treatment is carried out via a neutralization step in order to feed the water influenced in this way for subsequent treatment in the public wastewater system. As a rule, such plants do not have a purposefully equipped process stage for the treatment of PFAS. Under such marginal conditions, a noticeable reduction of the PFAS load cannot be assumed. b. Transport of the extinguishing water in silo vehicles to incineration plants: domestic waste incineration plants (850C), and hazardous waste or hightemperature incineration plants (1,100C). c. On-site treatment of firefighting water with activated carbon (GAC). Theoretically conceivable and already implemented in some practical cases is the use of large-volume activated carbon filters for the treatment of PFAS contaminated firefighting water. Depending on the respective PFAS contamination and the so-called organic and inorganic background contamination of the extinguishing water, it may not be possible to achieve the treatment objective at all or the costs resulting from the treatment may assume considerable dimensions. d. On-site treatment with the PerfluorAd process, in order to enable on-site treatment of PFAS -contaminated extinguishing water and also the risk-free use of activated carbon for such and other applications, the PerfluorAd process was developed, which significantly reduces the content of PFAS as a pretreatment stage, so that downstream process stages are significantly relieved and costs are reduced. Figure 8: Outline of the entry of firefighting water into the subsurface if no retention facilities are available taken from ( 2021). 224 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 9: On the left, a representation of a soil excavation after successful infiltration of extinguishing water into the subsoil. On the right the pump and treat procedure is shown ( 2021). Another stakeholder from Germany indicated that there is awareness about PFAScontamination and that the water is treated with adequate responsibility. In Bavaria (and also Germany in general) fire water-containment measures are in place that need to follow the "Lschwasser-Rckhalte-Richtlinie" (LRRL in english: "Extinguishing Water Retention Guideline", an English translation is not available). According to his knowledge the containment based on this guideline works (e.g. in industry plants), however, burning incidents involving large fires at facilities that are not covered by LRRi are more problematic. As an example, he named scrap tire storages (LfU-Gierig-Interview 2021). A stakeholder from the UK (LASTFIRE) informed that during/after fires the water run off containment has a lower priority than other concerns, at least this has been the case historically. However, adequate containment is possible in an industrial context but not followed in reality or not easy/cost effective to implement fully. In general, the containment of the water is not a problem for smaller fires, where the quantity of water is small, but can be for big fires. This is due to the fact, that the bunding might fail due to the high amount of water, or the bunding may not be sized to take account of the large amounts of water required. For the successful containment of PFAS-contaminated fire-run off waters the type and architecture of bunding areas is of highest importance and should be based on the amount of foam and water (e.g. in fire-fighting ponds) stored in the facility or the amount of fire and water required for a particular scenario (this information should be retained in the site emergency response plan). Today, the size of bunding area is typically calculated to have a holding capacity of 110 % of the largest tank, or where there are multiple tanks in a single bund 25 % of the total capacity of the tanks, whichever is the greater. There are primary, secondary, and tertiary bunding types. The primary containment is the tank itself. The secondary containment is the bund and the tertiary containment is beyond the bund but is designed to either contain a spill or direct the flow to a designed catchment area where it can be managed. Some of LASTFIREs members have taken adequate measures to prevent overflow of water, by having tertiary containment - often following reviews from the Buncefield incident. For jetty areas, the containment is even harder, and water would usually go nowadays directly to the sea. Another stakeholder from Germany explained that a complete containment of PFAScontaining run-off water is not in line with his real-life experience. In more detail, he explained that most of the run-off water is forfeit during the operations. Further, the stakeholder explained that there is almost always contamination of soil and water (DUS-ValentinInterview 2021). 225 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.4.1.1.3. PFAS-contents of PFAS-based AFFF and what can be measured Safety data sheets of PFAS-based AFFF often indicate a content of <5 % for fluorosurfactants. According to (Wood et al. 2020) this number can be further narrowed to a concentration range of 2-3 %. In practice, the foam gets further diluted in concentrations between 1 and 6 % (mostly in 1 %, 3 %, and 6 %). In a very recent publication by 2020) it is stated, that for waters with an undefined PFAS composition, as is to be assumed especially when using current AFFF foaming agents, an evaluation of the water load as well as the achievable cleaning results is not possible if only the quantifiable individual PFAS are evaluated. Held & Reinhard assume that AFFF foaming agents contain hundreds of precursor substances which include approx. 40 PFAS classes (Barzen-Hanson et al. 2017) and are highly complex and difficult to determine by classical analytical methods (Th. Held & Reinhard 2016). However, the precursors in such complex mixtures can be assessed by the Total Oxidisable Precursor (TOP) Assay (Mumtaz et al. 2019). In the same publication ( 2020), the composition of fire extinguishing water has been analysed using an exemplary product (not further specified) and various PFAS quantification techniques. In Figure 8 the results of this analysis are shown. Figure 10: PFAS-contents of a 1% AFFF Premix, measured using different analytical techniques. As shown in Figure 8, based on the analysis of 23 individual PFAS substances, as can be determined from parameter lists currently available on the market219, the foam had a total content of only 1.7 mg/l PFAS. It should be noted that precursor substances such as Capstone A and Capstone B are not yet included as parameters in the standard lists of environmental laboratories. In the example shown, these PFASs, which are often not yet quantifiable by measurement technology, have a concentration of 59 mg/l alone, i.e. these substances are almost 35 times higher in concentration than the individual PFAS compounds that are 219 DIN 38407e42 (status March 2011) served as the basis for the determination of PFASs. Methods using high-performance liquid chromatography and mass spectrometric detection (HPLC-MS/MS) after solid-liquid extraction (F42). With a simple matrix, determination limits of 10 ng/l can be achieved with this method for the individual congeners. However, in the case of complex matrix loads, the determination limits are often increased to several 100 ng/l. 226 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS quantified by PFAS standard analysis in laboratories. In order to address the total content of fluorine-organic sub- stance in complex contaminated waters, the organically bound fluorine was therefore used as an additional parameter for PFAS- contaminated waters that were contaminated due to exposure to AFFF foams. In the example in Fig. 3, the concentration of organically bound fluorine is 100 mg/l. Assuming that the average chain length and structure of the PFAS structures contained in current AFFF foams are significantly similar to the structure of the 6:2 FTS (H4PFOS), a hypothetical total PFAS concentration of 173 mg/l can be calculated. The hypothetical total concentration of PFAS leads, on the basis of the example shown to the conclusion, that the fluoro-organic substances in the water that cannot be detected as PFAS single substances can be a factor of 100 higher (or more) than actually measured PFAS by single substance analysis. This knowledge leads to the need to evaluate PFAS that are not known or quantifiable as single substances by sum parameters. However ( 2020), there is currently no normative standard for this. The above described results by 2020) show, that there is an ongoing uncertainty about PFAS-substances in AFFF. Based on stakeholder input in the context of this project, it can be stated that PFAS substances based on <C6-chemistry have never been used as an active ingredient for firefighting foams, as the chemistry is not suitable. <C6-substances are unintended by-products of the synthesis process (telomerization process) (FFFC-Interview 2021). 4.4.1.1.4. When does clean mean clean? Stakeholders have been indicated that there are uncertainties about what level are achievable when the success of cleaning procedures is to be judged. Concerning the wording: One stakeholder made the point that "decontamination" is to be distinguished from "cleaning". Cleaning needs to be done when one foam type is replaced by another (e.g. PFAS-based to PFAS-based). Concerned about remaining PFAS-levels, made the statement that by replacing the AFFF foam with a "truly" fluorine-free foam, there is already a positive effect for the environment. Also, if the PFAS-contaminated equipment of the fire departments would then be cleaned professionally in the course of this and the PFAS contamination remaining in the system were thus cleaned by 90, 95, possibly even 99 to 100%, this would be already a great accomplishment. But to prescribe a cleaning success of 100% bindingly, considers as not goal-leading ( Interview 2021). The same stakeholder also raised that fire extinguishing systems do not release foams permanently. Only at incidents or trips, PFAS-foams are emitted. This happens, in general, every 30-40 years per fire extinguishing system Interview 2021). Judged from the reported remaining PFAS-levels, it can be observed that a variety of PFAS-determination techniques is used. There are methods which concentrate on single PFAS-substances and also sum parameters are used. One stakeholder from Australia, indicated that setting cleanout standards for fixed foam systems and fire appliances was a concern in the Queensland foam Policy development stage as there were almost no precedents available. The primary issue was limiting cross-contamination of new foam put into existing systems from residues. From industry inputs and existing limits for foam concentrates we arrived at limits for PFOS + PFHxS in foams of 10mg/kg (UK and EU limits at the time) and 50mg/kg for other long-chain PFAS (by TOP Assay as F) as being practical and achievable for manufacturers and end users in systems and foam manufacture. It is not practical to set goals for cleanout based on wash-water concentrations as every system is different. Since then, experience has shown that much lower levels can be achieved. We have not set any recommended methods as each system is different in age, foam and components but we initially suggested that combinations of aggressive 227 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS surfactants, solvents such as glycol ethers and methanol may prove to be effective based on foam compositions and lab equipment cleaning techniques. Various fire engineer services have demonstrated that cleanout is achievable to those and lower levels depending on the circumstances, each has their own proprietary process and agents (Holmes 2020). 228 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.5. Non-destructive: Granular activated carbon (GAC) treatment 4.5.1.1.1. Background The treatment of PFAS-contaminated water (or PFAS-containing AFFF) with activated carbon is based on the adsorption of a molecule on the surface of the activated carbon. This is facilitated by van-der-Waals interactions between the activated carbon and the target molecule. As these interactions can occur between any two molecules a broad variety of compounds may be adsorbed, including some PFAS (mainly PFOS, this will be discussed in the technical performance sub chapter). This means that if a high concentration of other organic substances is present, the activated carbon becomes quickly fully loaded and unable to adsorb more molecules. As such the PFAS compete with other contaminants for the adsorption on the activated carbon surface. The carbon is typically supplied as powdered activated carbon or as granulated activated carbon (GAC) carbon (Analytik 2019; US-EPA 2020a). For the treatment with activated carbon the to-be-treated water is first filtered by a sand or multi-layered filter to filter out any non-solved contaminants and then sent through one or multiple activated carbon filters. By doing so the solved contaminants including PFAS adsorb to and saturate the surface of the activated carbon. If enough filters are installed in succession virtually all contaminants can be adsorbed out of the solution. The spent active carbon is either sent to reactivation or high temperature incineration. During reactivation high temperatures are used to thermally desorb the contaminants, which allows the reuse of the activated carbon. For this the spent carbon is heated up to 800 C for around 35 - 120 minutes. The conditions hereby range from a pyrolysis atmosphere (no oxygen) to a mild oxidative atmosphere (low oxygen) in order to restore the original carbon porestructure. An afterburner with temperatures between 880 - 1,316 C and a minimum residence time of 1 second is used to achieve a destruction rate of >99.99 % of the remaining contaminants. To what extent PFAS are destroyed under these conditions needs to be evaluated. Not all spent activated carbon can be reactivated. If the levels of organic halogens or metals is too high or the base carbon type is not suitable, a reactivation may not be possible. Alternatively, the activated carbon can also be incinerated via high temperature incineration. A reuse is therefore not possible (US-EPA 2020a) 4.5.1.1.2. Technical performance According to the ITRC-guideline, individual PFAS have different GAC loading capacities and corresponding breakthrough times (often defined as the number of bed volumes treated prior to detection in the effluent) (Eschauzier et al. 2012). GAC removal capacity for PFOS is greater than PFOA, but both can be effectively removed (McCleaf et al. 2017). In general, shorter chain PFAS have lower GAC loading capacities and faster breakthrough times but could be effectively treated if changeout frequency is increased. There are currently no published studies on the effectiveness of GAC in removing cationic, zwitterionic, and anionic precursor com- pounds; however, a recent theoretical study suggests some precursors are unlikely to be effectively removed by GAC ((Xiao et al. 2017) cited in (Ross et al. 2018). Furthermore, also the organic background of the water needs to be considered as this also lowers the efficacy as other organic substances can also bind to the GAC (Ross et al. 2018). Under optimal conditions i.e. using activated carbon with a high capacity potential, strongly absorbing PFAS, few competitive contaminants, low organic levels and a high concentration in the to-be-treated water loading rates of up to 0.1% can be achieved which corresponds to 1 g/kg of PFAS on the activated carbon. More realistic loading rates lie between 0,004 - 0,01 %(Analytik 2019; LANUV 2009; Maga et al. 2021a). According to Concawe, the US-EPA Health Advisory level for PFOS and PFOA (0.07 g/L) as well as the proposed EU drinking water threshold of 0.1 g/L for individual PFAS components (0.5 g/L for total PFAS) are achievable by activated carbon treatment, but may require the use of several beds in series (Concawe 2020). A PFAS removal rate of 99.9% has been 229 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS documented for a fire-extinguishing water treated by granular activated carbon. However, this elimination rate has been determined after a very short operation time. The maximum operation time until material exhaustion has not been reported. While higher influent concentrations would lead to higher loadings of PFAS onto the carbon, the presence of numerous co-contaminants may lead to a reduction in the loading due to competitive sorption. According to ITRC, most GAC full-scale treatment system case studies to date are based on treatment of PFOA and PFOS in the impacted drinking water sources. As such, limited information is available regarding the treatment of other PFAS. The full-scale drinking water systems demonstrate that PFOA and PFOS can be removed to below analytical detection limits until breakthrough occurs. Treatment of groundwater impacted with PFAS from an AFFF release area contaminated with PFAS such as fire training areas (FTAs) may require complex pre-treatment and more frequent change-outs (higher influent concentrations compared to influent for drinking water treatment systems) and higher operation and maintenance (O&M) costs. 4.5.1.1.3. Side products and emissions The adsorption removal mechanism of GAC is not expected to transform precursors (for example, telomer alcohols) to terminal PFAS as would be the case when using advanced oxidation/reduction technology (ITRC 2020). Emission may however arise when the GAC is reactivated or incinerated. For emissions from incineration see chapter 3.2.1. During the reactivation of GAC pyrolysis and gasification conditions are applied to restore the surface of the carbon. Hereby the carbon is heated to temperatures around 800 C under either a non-reactive (inert; no oxygen; pyrolysis) to mildly oxidising (steam and CO2; gasification) atmosphere. As the destruction of PFAS is achieved by completely oxidising all carbons of the PFAS molecule via the reaction with oxygen these processes may lead to different products. Especially under pyrolysis conditions where no oxygen is present small chain PFAS compounds and fluorinated gases may be formed. Typically, the facilities are equipped with afterburners operating between 885 - 1,316 C with a residence time of at least 1 seconds where all remaining contaminants are ought to be destroyed. As the formation of short chain fluorinated gases under the aforementioned conditions is likely it needs to be assessed whether the afterburner conditions can adequately destroy these compounds (US-EPA 2020b). According to (Ross et al. 2018) research indicates that some PFAAs can be destroyed on GAC surfaces at temperatures as low as 700 C during the reactivation process. Destruction of volatized PFAAs (in the air phase) requires 1,100 C; however, thermal reactivation kilns normally include after- burners for air pollution control, and these usually operate at temperatures above 1,100 C. Thus, a typical thermal reactivation process (800 C to 1,000 C reactivation temperature, plus an afterburner) seems to be well-suited for reactivating GAC that has exceeded its adsorption capacity for PFAAs. However, testing was not performed considering the wider range of PFASs, such as higher molecular weight (less volatile), polyfluorinated precursors reported to be associated with AFFF formulations. Data on whether these temperatures destroy all PFASs, including precursors potentially adsorbed to GAC, appears to be lacking. One stakeholder indicated that the activated carbon is mostly imported from China, used, and then re-activated in the EU. Reactivation is more profitable than buying virgin products. To his knowledge, the reactivation takes place at temperatures around 600 C, which could lead to incomplete destruction of PFAS and the formation of PFAS-side products. This could also lead to atmospheric deposition and contamination of soil and water ( Interview 2021). Availability across the EU Temporary and permanent GAC systems can be rapidly deployed and require minimal operator attention, if intensive pre-treatment is not needed (ITRC 2020). Currently, GAC is a 230 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS widely used water treatment technology for the removal of PFOS and PFOA, and, to a lesser extent, other PFAAs from water. Based on stakeholder input there is knowledge, that the activated carbon is mostly imported from China, used, and then re-activated in the EU. Reactivation is more profitable than buying virgin products ( Interview 2021). 4.5.1.1.4. Costs According to a recent report by the German Umweltbundesamt and Arcadis, the cost for the remediation can vary considerably (UBA 2020), for example from 0.40 - 2.30 /m in a pilot test. In another case, costs of < 0.06/m to 0.68 /m were found. Another study indicates the costs of sorption on activated carbon in the range of 0.24 /m (10 g/L PFAS in raw water) to 0.78 (100 g/L PFAS in raw water) (Q = 25 m/h). This includes electrical energy, maintenance, and activated carbon consumption. Based on these numbers an average cost of 1.25 per m PFAS-contaminated water is assumed, as calculated as the average of the respective highest reported cost value. 4.5.1.1.5. Additional information and available case studies Maga et al 2020 published a life cycle assessment comparing three treatment options for spent AFFF. In this study the authors compared the incineration, the treatment with granulated activated carbon and the treatment with PerfluorAd and subsequent activated carbon with one another. The focus was on the environmental impacts of the individual treatment methods e.g. greenhouse gas potential, resource depletion and emission of ionising radiation. In this study the treatment with GAC showed adequate results. GAC treatment emits large amounts of ionising radiation as most GAC is sourced from fossil coal deposits (Analytik 2019). Additionally GAC treatment can deplete the ozone layer as during the disposal of GAC many short chain side products may arise (Maga et al. 2021a). 4.6. Non-destructive: Ion exchange (IX) According to Concawe, no references were available for IEX treatment of fire-extinguishing waters or water with PFAS concentrations within the range of fire-extinguishing waters have been reported (Concawe 2020). However, as IX might be used as a secondary treatment after for example PerfluorAd, this method is shortly introduced as it is next to GAC the most established method. 4.6.1.1.1. Background According to ITRC, IX is an effective sorbent for other contaminants and has historically been used for a variety of water treatment applications (for example, nitrate, perchlorate, arsenic). To date, IX for PFAS removal from water is limited to ex situ applications (ITRC 2020). IX resin options for removal of PFAS include single-use and regenerable resins. Single-use resins are used until breakthrough occurs at a pre-established threshold and are then removed from the vessel and currently disposed of by high temperature incineration or by landfilling, where permitted. Regenerable resins are used until breakthrough but are then regenerated on site using a regenerant solution capable of returning the full exchange capacity to the resin. Temporary and permanent IX systems can be rapidly deployed. 231 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 11: PFAS flow diagram for adsorption filtration with IEX /taken from (Concawe 2020)). There is a variety of IX resins available at the market. According to Dupont, the polymer matrix of an ion exchange resin generally falls into two categories - gel or macroporous. A number of resins, both gel and macroporous type, developed for this market have similar chemical properties to allow for improved PFAS selectivity (Dupont 2020). 4.6.1.1.2. Technical performance According to the Concawe report and the therein cited literature, various anion exchangers have been identified with a higher adsorption capacity towards PFAS than activated carbons. The selective PFAS removal from contaminated waters by anion exchange works at both high PFAS concentrations of hundreds of mg/L as well as at low concentrations in the ng/L and g/L range. Similar to the adsorption onto activated carbon, the affinity of per- and polyfluoroalkyl sulfonates (PFSA) to ion exchangers is higher than those of per- and polyfluoroalkyl carboxylates (PFCA), and long-chain PFAS are absorbed preferably compared to short-chain PFAS. Treating groundwater, operation times up to 80,000 to 150,000 BV can be reached for the elimination of long-chain PFAS. However, retention of short-chain PFAS is lower and breakthrough starts at 10,000 to 30,000 BV. For ion exchange, the sorption kinetics for PFAS are relatively slow but it is still faster than adsorption on activated carbon. Fast sorption kinetics will result in a smaller filter geometry and therefore less investment costs. US and EU threshold value for PFOS and PFOA (0.07 to 0.1 g/L) are achievable using ion exchange resins. 4.6.1.1.3. Availability across the EU According to ITRC, Ion exchange technology has been used in the US since the late 1930s for common water treatment processes like softening, demineralization, and selective contaminant removal. The development and use of selective resins for PFAS removal is relatively new but already well established. As of 2019, a limited number of regenerable IX systems have been installed in full-scale applications after successful pilot testing. Collection of data on longer term treatment and on-site regeneration of the IX resin is ongoing at a case study site. Also, according to UBA, groundwater purification by means of ion exchangers is a common and widely used process. However, they have only rarely been used in Germany for the remediation of PFAS contamination. Accordingly, only limited experience is available from remediation on a technical scale. Due to the growing experience with this process, especially 232 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 12: Schematic overview of an Activated Carbon Plant (GAC) with PerfluorAd Pre-treatment Stage (taken from Highly PFAS-contaminated water can be for example fire run-off water or water from PFAS- related cleaning from technical equipment. According to the stakeholder, low PFAS-contaminated water is not the primary subject to PerfluorAd. Therefore, most groundwater contamination is not suited to be treated by PerfluorAd. GAC and techniques using ion exchanger (and combinations) are better suited. Also, AFFF-concentrates are not suited for PerfluorAd and would, theoretically, need to be diluted. The concentration of PFAS- and non-fluoride organic surfactants would be too high. Direct Incineration is the preferred option ( Interview 2021). PerfluorAd changes the solution equilibrium of PFAS in water. The reaction modes are precipitation and flocculation, mainly based on ion ionic interaction. The reaction is non- destructive meaning that the chemical composition of the PFAS substance is not changed ( Interview 2021). In addition to PFAS, PerfluorAd also removes other non- fluorinated surfactants. Those are used together with PFAS-surfactants in AFFF-products ( Interview 2021). For this a cationic compound mix consisting of different di- or triethanolamine quats (TEA) based vegetable fatty acids is added to the PFAS-containing water These fatty acids have the advantage of being biodegradable and synthesised from sustainable sources(Maga et al. 2021b) . The charged PFAS molecules interact with the positively charged "head" of the ethanolamine quats and precipitate (see Figure 11). The combination ratio thereby is not always 1:1 (Maga et al. 2021b). 234 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 13: The interaction between the PFAS molecule (below shown for the example of 6:2 FTS) and the added cation (taken from Maga et al 2020) Depending on the PFAS-concentration precipitant is added and is as such scalable. After precipitation the precipitated flakes can be filtered out (sand filter) and sent to high temperature incineration. 4.7.1.1.2. Technical performance In general, the removal efficiency of PerfluorAd is depending on the chain length and the polarity of the PFAS. The long-chain sulfonic acids (PFSAs) show the best removal efficacy. The same effect is also observed when using GAC ( Interview 2021). According to the the treatment with PerfluorAd can effectively remove a wide range of PFAS with an efficiency of up to 99.3 % for PFSA and PFCA. Other PFAS such as Capstone A and B220 only reach removal efficiency of 81 %. Substances like 6:2 fluorotelomer sulfonate (6:2 FTS) can reach a removal efficiency of 97 % and other non-PFAS surfactants can even be removed with up to 99,8 % efficiency 2020). The dosage ranges from 25 mg/L to 2 g/L and can be optimised for different PFAS concentrations and the water matrix to obtain higher elimination rates. In Figure 12 residual concentration total PFASs [mg/l] and right) elimination rate total PFASs [%] for 1% AFFF premix after addition of PerfluorAd are highlighted. Figure 26: left) Residual concentration total PFASs [mg/l] and right) elimination rate total PFASs [%] for 1% AFFF premix after addition of PerfluorAd (taken from ( 2020)) Principally, it is designed for treating water containing PFAS concentrations greater than 0.3 g/L (Ross et al. 2018). PerfluorAd is used as the first PFAS treatment step within a treatment train. Thus, this process is not intended to achieve final target threshold values (e. g. 0.1 g/L) as it is recognised that a further polishing step is required (Ross et al. 2018). The added PerfluorAd is specific for charged molecules so that in a recent experiment with diluted AFFF only 1.1 % of the dissolved organic carbon was precipitated. In this experiment 99.3 % of the total PFAS (23 substance) could be removed with the PerfluorAd treatment. 80.5 % of the Capstones and 87 % of the organically bound fluorine (precursors) were also removed with this process ( 2020). In Figure 13 elimination rates for different parameters [%] at an optimal dosing rate of 2.0 g/l are shown. 220According to UBA, The AFFF fire-extinguishing foams of products frequently used in Germany contain, in addition to some perfluorocarboxylic and sulfonic acids, the compound 6:2 FTS in low concentrations and, above all, in high proportions the two betaines (CAS: 80475-32-7 and 34455-293). 235 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS performed the best of the analysed treatment methods in nearly all investigated impact categories. Remediation of PFAS-contaminated groundwater under the Nuremberg Airport Fire Department's firefighting training area: the groundwater underneath the firefighting training area at Nuremberg Airport has been contaminated by PFAS due to the use of fluorine-containing firefighting agents over many years. A mobile groundwater remediation system based on the PerfluorAd principle was made available, thereby remediating the groundwater. The initial PFAS concentration in the groundwater was more than 600 g/l for the sum of the PFAS. With the PerfluorAd treatment alone, the PFAS load is reduced to 41 g/l (i.e., by 93.5%). After the final activated carbon stage (GAC for granulated activated carbon), PFAS contamination is no longer measurable. 4.8. Non-destructive: Foam fractionation and ozofractionation 4.8.1.1.1. Background Foam fractionation and ozofractionation are technologies that take advantage of the foamforming properties of PFAS. The process selectively separates PFAS from water by injecting compressed air (foam fractionation) or ozone (ozone fractionation) into the water in the form of air bubbles. PFAS surfactants adhere to the bubble walls and are thus transported to the surface. The PFAS-enriched foam is collected at the water surface for further destructionbased treatment. The treated water typically goes through a polishing step (e.g. GAC)(Concawe 2020). In the case of ozofractionation, precursors (also PFAS) are transformed to the perfluoroalkyl carboxylic (PFAA) and sulfonic acids (PFSA). Those PFAS remain in the system and are concentrated and discharged in the gas bubbles. Ozone can also promote the degradation of accompanying organic contaminants. Due to the small size of the gas bubbles (diameter < 200m), the total mass of the ozone bubbles has a large gas-water interface. At the surface of the water phase in the reactor, the PFAS are therefore concentrated in a small, separable volume. Figure 16: Illustrative Concept of foam fractionation (taken from (UBA 2020)) 4.8.1.1.2. Technical performance On a technical scale (Figure 7), the ozone fractionation consists of several reactors connected in series with continuous flow, into which ozone is introduced as bubbles. The PFAS concentrate as highly PFAS-contaminated foam on the liquid surface of the reactors. From the surface the bubbles get extracted via vacuum, further concentrated and can be fed to a further destructive treatment. The volume of the concentrate is 0.5 - 2 % of the inflow volume. According to literature the ozofractionation process alone mostly cannot achieve required PFAS concentrations and a supplementary process stage is needed. The gas phase is released into the atmosphere via an activated carbon absorber. As a rule, the last process stage of the water phase is an activated carbon absorber, with which the residues of the PFAS that have not yet been removed can be retained in order to achieve the required discharge values. If impurities are present, the process can be extended by further process stages if required. 237 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 17: Ozofractionation process concept (taken from (UBA 2020)). For long-chain PFAS such as PFOS and PFOA, a purification level of 99.9% has been achieved (Evocra 2017). For the ozofractionation stages alone, a purification level of >98.7 % was always achieved. The short-chain PFAS can be removed better with ozone than with air (Ross et al. 2018). The results further indicate that, for PFAS concentration levels below 0.3 g/L, high elimination down to a few ng/L could still be achieved (Evocra 2017). Similar to the precipitation with PerfluorAd, ozonofraction has an economic advantage at very high PFAS concentrations (which would be the case in PFAS-contaminated fire run off water and water from cleaning processes). The process is not only suitable for the treatment of water, but also for sludge with a solids content of up to 20%. The fractionation reactors separate the liquid from the solid phase. Small particles get into the foam concentrate and are removed with it. Coarse particles sediment at the bottom of the reactors and are removed there. Unlike many other processes, the degradation of an accompanying organic contamination does not significantly affect the PFAS removal level. The disadvantage is that a waste product (PFAS zone foam concentrate) is produced which must be disposed of separately (UBA 2020). Based on desktop research, it seems that the ozone is introduced to the reaction by adding of "Arcadis Solvent V171". The liquid has the following hazard statements: H227 (combustible liquid), H319 (causes serious eye irritation), H336 (may cause drowsiness or dizziness) and AUH019 (May form explosive peroxides). AUH019 is an Australian-specific H-statement and equals the European EUH019 (also may form explosive peroxides). Figure 18: Picture showing the GHS hazard statements of Arcadis solvent V171 (taken from a presentation of Arcadis at NEWEA, 2019 see here). 238 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Foam fractionation uses compressed air and is commercialized by the Australian company OPEC systems, allowing a continuous on-site treatment process in a containerized system. The treatment system is called Surface Active Foam Fractionation (SAFF). The operation mode of the system can be adjusted to manage a broad range of total detectable PFAS influent concentrations (0.1 to 100,000 g/L). The residence time per reactor vessel ranges from 5 to 30 minutes. PFAS-enriched foam is removed with a vacuum extraction system (Concawe 2020). According to the Concawe report and therein cited literature, for both methods , depending on influent concentrations, the US-EPA Health Advisory levels for PFOS and PFOA (0.07 g/L) as well as the proposed EU drinking water threshold of 0.1 g/L for individual PFAS compounds (0.5 g/L for total PFAS) are achievable without polishing. However, bboth technologies usually include a final polishing step, resulting in removal efficiencies of 99.9% to 99.99%. Very high influent concentrations might be managed via a multi-stage fractionation process (Concawe 2020). 4.8.1.1.3. Side products and emission Foam fractionation and ozofractionation are non-destructive techniques. In the case of ozofractionation, the PFAS-ozone bubbles are drawn off (vacuum extraction) and further concentrated and can be fed to a further destructive treatment. The volume of the concentrate is 0.5 - 2% of the inflow volume (UBA 2020). 4.8.1.1.4. Availability across the EU According to UBA/Arcadis, the ozofractionation process has already been tested on a technical scale in Australia. According to the available documentation, the process appears to be ready for the market. However, as is the case with most newer technologies, there is a lack of supplier-independent studies to verify its effectiveness. The supplier in Australia is a company called Evocra221, which signed a strategic exclusive agreement in 2019 with Arcadis. Based on research undertaken in this project, it seems that Arcadis promotes nowadays the usage of a V171 cleaning agent, which is most likely the same substances and related systems as the Evocra process. Foam fractionation is not available on a technical scale. However, limited field trials show promising results (OPEC-Systems 2020). 4.8.1.1.5. Costs Ozofractionation is a relatively complex technology whose operating costs are significantly higher than those of alternative market-ready technologies (e.g. GAC) but this cannot be assessed due to lacking data (UBA 2020). 4.8.1.1.6. Additional information and available case studies The technique has been used in several cases in Australia and one in the UK, this involved (based on the results of the desktop search within this project): A large-scale implementation of ozofractionation at an airport in Australia using a NF unit for polishing to treat PFAS affected surface water and wastewater achieved a removal efficiency of 97% for the sum of 28 PFAS with inlet concentrations of 100 to 5,400 g/L222. Water remediation at a fire training site223 Containing of 22,000 liters of escaped PFAS contaminated waste from a failed deluge system within an airport hangar224. 221 See Evocras internet site here, last accessed 01.04.2021 222 See site here, last accessed 01.04.2021 223 See presentation here, last accessed 01.04.2021 224 See presentation here, last accessed 01.04.2021 239 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Contamination stemming from an airport in the UK (Guernsey island)225 US-Department of Defence (DoD) concerning the "Demonstration and Validation of Environmentally Sustainable Methods to Effectively Remove PFAS from Fire Suppression Systems" 4.9. Destructive approaches 4.9.1. Incineration The details of PFAS-based incineration are explained in detail in chapter 3.2.1 (hazardous waste incinerators) and 3.2.2 (cement kilns). For the incineration of PFAS-contaminated water, the same information applies as in those chapters. However, as the concentration of PFAS-contaminated fire run off water and cleaning water are considerably lower, literature indicates that in some cases non-destructive techniques are used in order to lower the to be incinerated volume and related costs. 4.10. Conclusion on the Disposal of PFAS-contaminated (fire run off and cleaning) water The following conclusions can be made for available disposal options for PFAS-contaminated (fire run off and cleaning) water: Fire run off and cleaning water are highly PFAS-contaminated compared to for example groundwater contaminations. Based on this not all available remediation techniques for groundwater can be used also for run off and cleaning water. GAC can be also used for all PFAS-contaminated for run off and cleaning water. However, the efficiency is lower for PFAAs (carboxylic acid) in general and short chain PFAS. For other PFAS (e.g. zwitterionic) no data is available. One stakeholder brought up that when GAC is reactivated (using 800 C) PFAS could be emitted ( Interview 2021). Ion exchange (IX) is generally suited for PFAS-contaminated run off and cleaning water. However, no caste studies are available. Based on the type of PFAS various IXmatrices are available. IX is believed to be 4 times more expensive that GAC, when only the material is considered. According to UBA, the total costs for ion exchangers compete with the costs for the sorption of the PFAS on activated carbon. Even if the activated carbon process is less efficient and requires more sorption material, in the end it could be cheaper (UBA 2020). However, there are no actual costs cited. GAC and IX are generally based on column beds to which PFAS absorb. To achieve certain PFAS-levels several beds in series must be used. With both techniques proposed EU drinking water threshold of 0.1 g/L (0.001 ppm) for individual PFAS components (0.5 g/L for total PFAS) are achievable but may require the use of several beds in series. The material cost for GAC is around 0,41 - 3,68 /kg. According to a recent report by the German Umweltbundesamt and Arcadis, the cost for the remediation can vary considerably (UBA 2020), for example from 0.40 - 2.30 /m in a pilot test. In another case, costs of < 0.06/m to 0.68 /m were found. Another study indicates the costs of sorption on activated carbon in the range of 0.24 /m (10 g/L PFAS in raw water) to 0.78 (100 g/L PFAS in raw water) (Q = 25 m/h). This includes electrical energy, maintenance, and activated carbon consumption. Based on these numbers an average cost of 0,85 per m PFAS-contaminated water is assumed, as calculated as the average of the respective cost values. 225225 See presentation here, last accessed 01.04.2021 240 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS For IX, material cost is about 12 /kg and treatment costs for long-chain PFAS of 0.05 to 0.1 /m and for short-chain PFAS of 0.25 to 0.8 /m respectively. Based on these numbers an average cost of 0.45 per m PFAS-contaminated water (for both longand short-chain PFAS) is assumed. To minimize the load (and therefore costs) of GAC/IX, precipitating agents like PerfluorAd can be used. The active ingredient changes the solubility of PFAS. PFASPerfluorAd sludge can be incinerated. The water then is then further treated with GAC/IX (treatment train). For the entire PerfluorAd/activated carbon system, operating costs (depending on the activated carbon used) amounted to < 0.055 - 0.68 per m of treated water, of which approx. 0.04 /m is attributable to the PerfluorAd requirement. Ozonofraction uses the fact that PFAS remain the air-water interface and creates ozone-bubbles which are considerably smaller than regular air-bubbles. Bubbles then can be physically removed. The water then is then further treated with GAC/IX (treatment train). For PFAS concentration levels below 0.3 g/L, high elimination down to a few ng/L could still be achieved. No information is available for the costs of this technique, however, ozonofraction is a complex technology whose operating costs are significantly higher than those of alternative market-ready technologies (e.g. GAC), but this cannot be assessed due to lacking data. PFAS-contaminated fire run off and cleaning water can also be directly subjected to incineration. The cost for the disposal of 1 liter of PFAS-based AFFF are currently in the range of 0.2-2 /l (around 200-2000 /m3), it can be assumed that the same costs apply to fire runoff water. Based on the available data, the direct incineration of PFAS-contaminated run-off water would be the most expensive disposal alternative (200-2000 /m3). According to (UBA 2020)GAC and IX are comparable in costs (although material costs differ). For GAC an average cost of 0,85 per m PFAS-contaminated water is assumed (three projects considered). For IX an average cost of 0.45 per m PFAS-contaminated water (for both long- and short-chain PFAS) is assumed. Based on available data, the combination of PerfluorAd and GAC is the cheapest technique with an average reported cost of < 0.055 - 0.68 per m of treated water, of which approx. 0.04 /m is attributable to the PerfluorAd requirement. The cost depends on the activated carbon used. Based on these numbers an average cost of 0.37 per m PFAScontaminated water (for both long- and short-chain PFAS) is assumed. 241 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 244 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Cost Low estimate Best estimate Replacement of foam stocks (only RMO 2) 320 million 1.0 billion Total one-off costs Annualised total one-off costs Annual costs Additional volumes of alternative foams 100 million (RMO 1) 630 million (RMO 2) 9.0 million (RMO 1) 57 million (RMO 2) 21 mill on (RMO 1) several millions (RMO 2) 1 billion (RMO 1) 2.3 billion (RMO 2) 90 million (RMO 1) 210 million (RMO 2) 27 million (RMO 1) 10 million (RMO 2) (assumed midpoint between low and high estimate) Disposal of expired foams Clean-up - several millions -10s of millions -1 million (assumed midpoint between low and high estimate) -10 million (assumed midpoint between low and high estimate) Treatment of fire-water run- -10s of millions - several mill ons High estimate 2.0 bill on 1.5 billion (RMO 1) 3.9 billion (RMO 2) 130 million (RMO 1) 350 million (RMO 2) Notes available to quantify this. Range based on central estimate pr ces per tonne of PFAS-based foam and alternatives, and likely range of tonnage of alternatives to be purchased. When cons dering full range of possible prices per tonne, the range of total replacement cost could be - 60m 8.3bn. As per the notes above, the possible range could be even wider (low estimates 530m lower, high estimates 10.7bn higher). As per the notes above, the possible range could be even wider (low estimates 48m lower, high estimates 960m higher). 30 million (RMO 1) 20 million (RMO 2) (assumed value <30 million) -100,000 Under RMO 2, all PFAS foam stocks are replaced with new alternative foams at the beginning of the assessment per od (already covered under the one-off cost replacement of foam stocks" above). These new foams would not expire within the assessment per od, but an unknown share would be used and still need to be replaced again with new foams, thus incurring the costs associated with additional volumes again. Therefore, this cost is lower under RMO 2 but t cannot be quantified by exactly how much. -1 million (assumed value close to 0) 0 High estimate based on the assumption that at least in some cases, savings from reduced clean-up will be incurred, so total savings will be somewhat larger than 0. 248 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Cost off per inc dent (annual unknown) Total annual costs Low estimate - 10s of millions (i.e. a benefit) Best estimate (assumed midpoint between low and high estimate) several millions (RMO 1) High estimate Notes 29 million (RMO 1) 19 million (RMO 2) Total annual costs + annualised one-off costs - 10s of millions (RMO 1) ~0 (RMO 2) - several millions (i.e. a benefit) (RMO 2) ~100 million (RMO 1) ~200 million (RMO 2) Results rounded to two significant figures. 160 million (RMO 1) 370 million (RMO 2) As per the notes above, the possible range could be even wider (low estimates 48m lower, high estimates 960m higher). As a way of comparison, the Firefighting Foam Coalition indicated in their comment to the PFHxA restriction proposal that taking into account the loss of foam sales and the cost for procurement and disposal of foam agent and equipment, they estimate that the restriction as proposed would cost EU foam manufacturers and users more than 200 million230. Cost effectiveness Based on the above, as a central estimate, it is calculated that the cost effectiveness could be around 850 (RMO 1) to 1,700 (RMO 2) per kg of annual reduction of PFAS emissions. This could range from savings in the tens of Euros per kg to costs around 10,000 per kg. Table E.19 Estimate of cost-effectiveness of the reduction of PFAS emissions from firefighting foams Low estimate Best estimate High estimate Total emission reduction (kg) 234,000 kg 117,000 kg 36,000 kg Total cost () - 10s of millions (RMO 1) 100 million (RMO 1) 160 million (RMO 1) ~0 (RMO 2) 200 million (RMO 2) 370 million (RMO 2) Costeffectiveness - 10s /kg (RMO 1) 0/kg (RMO 2) 850/kg (RMO 1) ~1,700/kg (RMO 2) 4,600/kg (RMO 1) 10,000/kg (RMO 2) Results rounded to two significant figures and reflect the likely range. However, as noted in the prev ous table, the range could possibly be even w der, from - 10s /kg (both Scenarios) to 31,000/kg (RMO 1) and 37,000/kg (RMO 2). 230 Comment 3010 in Part 2 document available at: https://echa.europa.eu/restrictions-underconsideration/-/substance-rev/25419/term 249 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS E.8.2 Stakeholder input on transition periods Several users have provided input on manageable transition periods: One stakeholder claimed that a transition time of 10 years would be needed for the switch in the O&G / petrochemicals sector. Another stakeholder from the same sector cited 5-10 years, in order to minimise and spread the costs to change foam and re-build, or re-place fire extinguishing systems or equipment, but they would like to keep PFAS stocks in case of a big fire incident. As reported in the case study in Section E.2.4.2, Equinor took around 8 years to transition to fluorine-free foams. An industrial end user under consideration of discussions with some representatives from aviation industry groups and municipal users has developed a detailed draft proposed timeline covering a range of tasks required for a full transition to fluorine-free foams (across all sectors). The full timeline is provided in Annex 4, but key milestones suggested are (years from formal start of transition and introduction of legislation): No more PFAS foam use in training: Immediately No more PFAS foam use in systems testing: 4 years No more PFAS foams used for small incidents: 4 years Completion of transition: 10 years. The additional 6 years from the previous steps is largely driven by further replacement and disposal of stocks of legacy foam231, as well as the need for further development of fluorine-free foams by manufacturers. A stakeholder from the aerospace and defence sector suggested the system change to enable use of non-PFAS foam could be introduced at time of major refit, which typically occurs every 6 -12 years. On the other hand, the US Fiscal Year 2020 National Defense Authorization Act (NDAA) requires a phase-out of PFAS-containing firefighting foam in the US military (except on ships) by October 2024, i.e. within 4 years. Several stakeholders across different sectors stated at the workshop or in response to the written consultation that 3-6 years may be sufficient. One stakeholder suggested different transition periods for different uses. They explained municipal fire brigades should be able to transition quicker than operators of fixed installations for example. They argued that the use of fluorine free foam for tank fire fighting needs further testing and therefore more time. The PFHxA proposed restriction foresees the following transition periods: Concentrated fire-fighting foam mixtures placed on the market until 18 months after the entry into force of the restrictions can be used in the production of other firefighting foam mixtures until 5 years after the entry into force, except for use of fire-fighting foam for training and (if not 100% contained) testing. There is also an exception proposed for concentrated fire-fighting foam mixtures for certain defence applications until a successful transition to alternatives can be achieved, and for concentrated fire-fighting foam mixtures for cases of class B 231 Note that this does not necessarily imply that no more PFAS based foams are purchased during that period. 252 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Impacts Economic Social Health/Environmental W der economic implications Speed of fire suppress on may be slower and applicat on of foams may be less flexible and less easy to use, according to some stakeholders. This has not been shown to be generally the case and resulting health/safety impacts could not be quantified. d. Use patterns to achieve comparable/acceptable performance using alternatives Between no change in volume and up to a maximum of 100% additional foam required (additional cost cons dered in e. below). None identified. In sprinkler appl cations, special sprinkler nozzles have to be installed (cost not quantified). More than one foam may need to be stocked by users to cover different flammable liquids, with logistical, training and safety implications for users. None identified. e. Econom c feasibility of alternatives More than one foam may need to be stocked by users to cover different flammable liqu ds, with logistical, training and safety implications for users. For both Scenarios: None identified. None identified. None identified. 255 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Impacts Economic Social Most likely there is no significant pr ce difference (per litre) between PFAS-based foams and alternatives, but up to 100% more volume may be required (central estimate 50%) to achieve desired performance. This would lead to costs around 27m per year (EU total, central estimate) Potential additional economic costs for trans tioning may include testing costs (not quantified), storage costs, (not quantified) costs from technical changes to delivery systems (e.g. 5-200 pre nozzle or around 2,700 for a mobile foam unit, but generally conceived as manageable), and regulatory approvals (not quantified). Potential savings may include lower foam disposal costs at expiry date (likely order of magnitude 100,000 to several million) lower fire-water disposal costs (covered under g. Remediation and clean-up), avoided cross contamination of waste, reduced regulatory requirements and reduced PPE requirements (not quantified). Additionally for RMO 2: Costs for existing stock of PFAS-based foams (estimated 210,000-435,000 tonnes) will have to be written off (and new stocks Health/Environmental W der economic implications 256 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Impacts Economic Social f. Environmental/health - impacts of alternatives purchased causing an addit onal cost (central estimate 1.0bn) over the baseline. None dentified. None identified. g. Remediat on and clean-up Potential risk of PFAS contamination could be eliminated, which could save up to around 100 million remediat on costs per s te. Depending on the extent of containment and immediate clean-up, the number of relevant s tes is likely low, but overall savings could still be in the order of magnitude of 100s of millions to billions More informat on on the total number of s tes and real-world implementation and effectiveness of best pract ces would be required to be more precise. None identified. Health/Environmental W der economic implications Based on the assessed substances, nonfluorinated alternatives are of lower environmental concern, primarily due to greater biodegradat on. A quant tative comparison of risk under each scenario was not possible w th the available data. Potential trade-off between remediation cost and remaining PFAS contamination. None identified. None identified. h. Availability of alternatives Treatment of fire-water run-off and shortterm clean-up largely driven by other contents of fire-water run-off and cost saving estimates are very uncertain. Run-off treatment savings could be 0.7 per litre (range ca 0-11) or 0 to 10s of millions per incident, and clean-up cost savings up to 10s of mill ons in total. Range of alternatives and capacity to increase product on likely available. No significant None identified. None expected in RMO 1. The risk of supply supply-shortages is higher in RMO 2 (depending on timescales of a restr ction), None identified. 257 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Impacts Economic Social supply shortages or add tional costs expected in RMO 1. RMO 2 may result in a more sudden and potentially significantly larger demand for fluorine-free foams to replace existing stocks of PFAS-based foam. This heightens the risk of a shortfall in supply, - depending on the timescales of any restriction. As the largest single use, and with comparatively low current fluorine free sales volumes, the risks of supply constraints may be greater in the chem cals and petrochemical sectors. Health/Environmental which could potentially lead to addit onal firesafety risks. W der economic implications i. Other impacts None dentified. There is potential for employment impacts but significant impact is deemed unlikely and any net effect at the EU level would be negligible. None identified. Under RMO 2, a potential shortfall in supply - driven by a one off need for stock replacement may impact imports of fluorine free foam from outside the EU. 258 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Impacts Economic Social j. Emissions from disposal of legacy foams None dentified. None identified. k. Techn cal feasibility / availabil ty of disposal options If the transit on period is short, there is the demand for disposal facilities may outstrip supply, leading to potential add tional costs and potential for emissions. None identified. l. Costs of disposal Total EU costs estimated at up to 320 million (range up to 60m-4.8bn) depending on the method used (with implications on effectiveness, see Health/Environmental) and the share of foams that would have reached expiry date w thout use. None identified. Additional transport, storage and labour costs may be incurred (not quantified). Health/Environmental W der economic implications Overall, there are unlikely to be any signif cant macroeconomic impacts from the result of either RMO 1 or RMO 2. High temperature incineration has been dentified as main disposal method. There are potential hazards (emissions of hydrogen fluoride and PFAS), but further research is needed to identify and quantify the emissions produced from the incineration of PFAS. None identified. Trade-off between temperature of incineration (w th lower capacity and higher costs) and effectiveness of PFAS destruction. None identified. If the transition period is short, there is the demand for disposal facilities may outstrip supply, leading to potential addit onal costs and potential for emiss ons. Trade-off between temperature of incineration (w th lower capacity and higher costs) and effectiveness of PFAS destruction. None identified. 259 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Annex F: Assumptions, uncertainties and sensitivities The above conclusions are subject to a range of assumptions and uncertainties. Assumptions have been made based on the results of other sections and are discussed in more detail within those sections. However, the assumptions and uncertainties that could most significantly affect the results are discussed briefly below: Environmental/health benefits of the reduction of PFAS emissions could not be quantified, primarily due to a lack of knowledge about the hazards of PFAS. The estimated emissions of PFAS and hazards of the constituents of alternatives are also subject to a range of uncertain assumptions. Hence, costs and benefits could not be directly compared; Cost of transition are subject to uncertain assumptions about price difference between foams and the quantity of alternative foams needed to achieve the desired performance. Which and how much alternative foam is needed to achieve the desired performance varies on a case by case basis. It has been judged most likely that there is no significant price difference (per litre) between PFAs-based foams and alternatives, and assumed that 50% additional volume of alternatives is needed. If a more/less expensive alternative foam or larger/smaller quantities would be needed to achieve the desired performance, this would increase/decrease the costs of the restriction. Savings related to the transition are sensitive to assumptions about the amounts of foam that would reach their expiry date without use under the baseline; Costs of cleaning and technical changes or replacement of equipment are very casespecific and could largely not be quantified with the available data. This means that the quantified costs of both scenarios are underestimates; It should be noted that there was a divergence in the stakeholder input about technical feasibility of alternatives. A few stakeholders have voiced concerns over the potentially reduced fire safety, at least in specific applications. This means there is a risk of additional health, safety and economic (fire damage) impacts; however our analysis has concluded that they are not the most likely outcome and that LAST are the main application for which there is still further testing required; It is uncertain to what extent current practices involving the use of PFAS-based fire-fighting foams already manage to eliminate the need for significant new remediation requirements under the baseline. This is because most experiences with PFAS remediation relate to legacy contamination resulting from historical emissions before current measures (e.g. containment and clean-up after use) were widely implemented. However, stakeholder input suggests that such measures are likely not 100% implemented or effective. Furthermore the incremental costs of addressing PFAS contamination in short-term clean-up is difficult to separate from the wider clean-up costs involved after fire incidents. In addition, there is a lack of data about the number of sites that use PFAS-based foams in sufficient quantities to potentially require clean-up or large scale remediation. Therefore remediation savings from the transition to fluorine-free alternatives are very uncertain and only illustrative estimates of the potential order of magnitude of such benefits were provided; and There is a wide range (60-4,800 million, with best estimate 320 million) in the potential costs of disposal of legacy foams in RMO 2, which largely depends on the disposal method used. This is due to uncertainty about the amounts of foam that would reach their expiry date without use under the baseline and the effectiveness of PFAS-destruction at different 260 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS incineration temperatures and times. There is also uncertainty about the potential emissions and therefore associated environmental/health risks. 261 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Annex G: Stakeholder information Introduction At the beginning of this project it was recognised that effective engagement with key stakeholders from across the fire-fighting foam sector, particularly the manufacturers and users of the foams, would be critically important in the data collection process. The stakeholders consultation cover all the relevant sectors and backgrounds across the firefighting foam supply chain, as well as regulators, researchers and special interest groups. The consultation therefore aimed to target the following stakeholders: Foam manufacturers / suppliers; Users of foams in major sectors (including airports, oil and gas, chemical plants, ports, railways); Key trade associations; International organisations; National-level authorities and agencies; Academics and R+D (especially those involved in developing alternative foam products); and Key NGOs and interest groups. The consultants Wood, Ramboll and COWI mapped stakeholders identified so far, indicating the best means of consulting each one of them: e.g. advisory group, questionnaire, one-toone consultation, workshop, etc. An initial list of stakeholders was provided by the consultants and a finalised list was agreed with the Commission and ECHA prior to commencing the consultation activities. Approach The agreed approach to collecting key information from the main categories of expert stakeholders (detailed above) was to carry out a consultation through a combination of i) scoping interviews, ii) a targeted stakeholder questionnaire, and iii) a stakeholder workshop. The approach to carrying out these consultation activities is outlined in the following sections. Scoping interviews To inform the main data collection steps of the project (the stakeholder questionnaire and workshop) a series of initial scoping interviews was organised with a selected small number of key stakeholders. The purpose of the interviews was to: I. Introduce and discuss the aims and scope of the project with key experts; II. Identify where the key data gaps were in relation to the objectives of the project(s); and III. Identify other key stakeholders in this sector to target in the next stages of the consultation. 262 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS The stakeholders involved with the scoping interviews were: Eurofeu; Fire Fighting Foams Coalition; Copenhagen Airport; Heathrow Airport; LASTFIRE project; and IPEN. An interview template was developed, and shared with the interviewees ahead of the call, to guide the conversation more effectively and efficiently. Teleconference interviews of 30-60 minutes were held with each stakeholder. During the call, brief notes of the key discussion points were made. Since the purpose of these scoping interviews, was as an introductory discussion, rather than an evidence gathering exercise as such, a limited amount of specific information about the use of fire-fighting foam products was gained. A number of key outcomes from these scoping interviews are highlighted below: All stakeholders interviewed expressed an interest in the project and agreed to participate in the consultation; In some cases, for example, for key industry associations, it was agreed they would coordinate joint industry responses, and stakeholders provided the details of additional stakeholders to contact, and/or agreed to forward the consultation on directly; Both industry, users and others (e.g. NGOs) commented on the increased move towards and the rapidly increasing market share of fluorine-free foams, and their increasingly better overall performance now than previously; It was re-emphasised that alternative foams are designed for very specific applications, requiring compliance with specific performance criteria, so the analysis of their technical and economic feasibility will be challenging as it requires assessment of each product individually; There is likely to be variation in the situations with regards to alternative foams in different sectors of use (e.g. between aviation and oil and gas sectors) and in different locations/countries (e.g. certain countries have switched to alternatives, others have not); and The potential for contamination of foams was raised, leading to the inclusion of specific questions in the survey about the level of PFAS as impurities in foam products (both PFASbased and fluorine-free). The scoping interviews were then used to better inform our approach to the following consultation steps, allowing the survey and workshop to be designed more systematically to address the key knowledge gaps and target the most relevant stakeholders. This also helped to identify additional stakeholders to include in the next consultation steps. 263 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Consultation questionnaire The main consultation activity conducted involved the development of a written questionnaire, based on an assessment of the required data needed to generate and/or complement the information already gathered under the separate Tasks under the two projects. It was agreed that the most appropriate format of the questionnaire would be a Word document that could be sent to targeted stakeholders directly via email, allowing the respondents to fill in relevant details and return the completed questionnaires. The questionnaire covered the following aspects: Introductory information; Background information on the respondent; Chemical identity, functionality of PFAS in fire-fighting foams; Alternatives to PFAS in fire-fighting foams; Foam use and environmental emissions; Potential restrictions on PFAS in fire-fighting foams; and Additional information. The full consultation questionnaire is provided in Appendix 1 of the underlying study233. Consultation questionnaire results A total of 33 written responses to the questionnaire were received234. Of the different types of stakeholder targeted, the most responses were from users/industry (11), with smaller numbers of responses from individual manufacturers (2), authorities/agencies (6), industry associations (2), NGOs (3) and `other' stakeholders (7) e.g. academic/testing/training professionals/technical consultant. It is noted that the responses from the users of foams cover all the main sectors of use the consultation aimed to cover (airports, oil refineries/storage, chemicals, petrochemicals, and rail). Responses to the consultation from a number of stakeholders also included the provision of previously published data or reports in addition to, or instead of, the questionnaire. This included published reports and analyses from national authorities235, research and testing information236, and special interest groups237 233 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 234 Correct as of July 18 2019. 235 KEMI (2015) Chem cal Analysis of Selected Fire-fighting Foams on the Swedish Market 2014 236 Published testing data, as provided by LASTFIRE: www.lastfire.co.uk/ 237 IPEN (2019) The Global PFAS Problem: Fluorine-Free Alternatives as Solutions, https://ipen.org/documents/global-pfas-problem-fluorine-free-alternatives-solutions 264 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Since the questions in the stakeholder questionnaire were designed to gather information that will best feed into the delivery of tasks under each of the two projects, the responses received have generated useful information in this context. In particular, we highlight the following aspects, where the consultation yielded useful information: Identifying some of the key foam products containing PFAS on the EU market, and nonPFAS alternatives actually used in key sectors; Identifying specific PFAS, precursors and impurities present in some foam products; The functionality of PFAS-containing foams useful to the major users of foams and reasons why products containing PFAS have not been fully replaced; Volumes of production and use, and unit price for a small number of individual products; Information on available alternatives, including specific products on the market in the EU, the type and sector of use, their availability, volumes of sale and use, their perceived technical feasibility and economic feasibility; Some details of fire-fighting foam use e.g. volumes, frequency; Some details of methods, regulations, and guidelines in place to prevent release to the environment; Some information on the methods/approach to disposal of individual foam products; Preliminary stakeholder opinions and feedback on different potential restriction options were provided; and Additional data, reports and other resources were provided by a number of stakeholders with their consultation response. For some sections, a number of information gaps, where the level of detail provided by respondents was less substantial, were identified. These data gaps helped to inform the approach to the organisation and format of the following stage of the consultation process, the stakeholder workshop, where these data gaps were explored further (see next section). Consultation workshop The final stage of the consultation involved the organisation of an expert stakeholder workshop. This was hosted by ECHA in Helsinki on 24 September 2019. The purpose of the workshop was to present, validate and seek feedback on the preliminary project findings; gather views on possible risk management options; and explore the feasibility of replacing PFAS-based foams with fluorine-free alternatives. Stakeholder views were sought during the workshop through a series of breakout groups on key topics which focused on specific questions designed to inform possible future regulatory activities. The workshop was attended by a total of 36 participants, including manufacturers, users from different sectors (airports, chemical plants, oil and gas), researchers, NGOs, national authorities, and remediation experts. The format of the workshop included: 265 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Introductions from DG Environment, ECHA and the study team; Presentation of initial results; Plenary discussion on study findings to date; Presentations from invited speakers; and Breakout session on remaining data gaps. The invited speakers, who presented at the workshop were from the following organisations: Eurofeu (industry perspective); Finavia Corporation (user perspective - airports); Total HSE (user perspective - oil and gas); and LASTFIRE project (testing and efficacy perspective). There were four breakout sessions for the workshop, each covering a specific set of questions, partly informed by the identified data gaps remaining from the consultation questionnaire and the other tasks relating to the two projects. The breakout sessions covered the following aspects: Different Risk Management Options; Essential uses and availability of alternatives; Remediation costs and technologies; and Current/ future market trends in PFAS-based and fluorine-free foams. The workshop report with more details about the set-up and results of the workshop is included in Appendix 2 of the underlying study238. Additional consultation and resources The stakeholder consultation and workshop also resulted in a number of stakeholders providing additional information to supplement their consultation responses. This additional information was used, where relevant, in each of the specific tasks. Following the consultation questionnaire and workshop, a number of specific areas were identified as needing additional data or clarification, for example on volumes of firefighting foams produced, marketed and used in the EU. Where these additional data needs were identified, the project team undertook direct consultation with specific stakeholders identified as being the best source of the required information. Contact was made with these stakeholders via email or telephone to discuss the remaining data needs and obtain the 238 Wood, Ramboll, COWI: "The use of PFAS and fluorine-free alternatives in fire-fighting foams - Final report". Report for the European Commission DG Environment and European Chemicals Agency (ECHA) under specific contracts No 07.0203/2018/791749/ENV.B.2 and ECHA/2018/561. 266 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS required data. This additional consultation has provided additional detail and clarifications relating to critical uses, volume of production and use in the EU, and experiences from previous transitions. 267 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.12. V171 by Arcadis 4.12.1. Background According to the JOIFF-article (which has been authored by Ian Ross and Storch from Arcadis), decontamination of firefighting and fire suppression equipment is essential to limit carryover of PFASs from old foam usage. Triple rinse with water is not sufficient and leads to a significant volume of decontamination water that requires treatment. Arcadis recommends using specialized biodegradable cleaning agents such as V171 to effectively remove PFAS residuals from fire suppression systems to limit future liabilities and cost associated with PFAS contaminating F3 foams as a result of inadequate decontamination (JOIFFF 2020). Arcadis has developed methods for PFAS decontamination of piping and tank systems including the use of a proprietary biodegradable cleaning agent, V171. These methods and the cleaning agent have been successfully applied in foam transition projects to remove PFASs from steel and PVC piping systems, stainless-steel concentrate tanks, and underground wastewater tanks (JOIFFF 2020). Also application in foam suppression systems, emergency response vehicles, and concrete sewer distribution systems are described (Anderson 2021). In chapter 5.5, the technical performance and other details of this technique has been described. The following information concentrate on the actual cleaning procedure. 4.12.2. Cleaning Procedure The initial PFAS cleanout project in 2017 used a sequential series of aqueous rinses, high- pH flushes and application of the cleaning agent as shown in Figure 17. presenting Sum of PFAS (28) TOP Assay results. The results demonstrated that water and high pH are ineffective for removal of PFASs from surfaces, as demonstrated by the relatively low concentrations of PFASs measured in these flush solutions. The elevated concentration in the cleaning agent demonstrated significantly greater PFAS mass removal even after multiple flushes of water and caustic solution. Further work to clean PFASs out of a 20- m foam concentrate tank was conducted, and results are presented in Figure 17. This application demonstrated that soaking with the cleaning agent, followed by high-pressure washing can be effective. The importance of using TOP assay for analysis of PFASs was revealed. 269 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 19: Sum of PFAS Concentrations during decontamination of AFFF- Impacted sewer system and of a 20-m concentrate tank 4.12.3. Remaining PFAS concentrations As shown above in the diagrams of Figure 17, the final water flush/rinse contained around 0.1 g/l PFAS as measured for the sum of 28 PFAS (according to TOP 4 g/l). 4.12.4. Costs There is no information available on costs for this technique. 4.12.5. Additional information and available case studies Available case studies has been already discussed in chapter 5.5 under "Additional information and available case studies". In addition, Arcadis claims that the technique has been successfully applied in foam transition projects to remove PFASs from steel and PVC piping systems, stainless-steel concentrate tanks, and underground wastewater tanks (JOIFFF 2020). Also application in foam suppression systems, emergency response vehicles, and concrete sewer distribution systems are described (Anderson 2021). For these projects, no documentation has been found via desktop search. One ongoing study for the DoD focuses on Fire Suppression Systems. US-Department of Defence (DoD) concerning the "Demonstration and Validation of Environmentally Sustainable Methods to Effectively Remove PFAS from Fire Suppression Systems". 4.13. Cleaning procedure with PerfluorAd by 4.13.1. Background Detailed information on the background of this technique has been already given in chapter 5.2. 270 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Cleaning Procedure Cleaning of PFAS Contaminated Fire Fighting Trucks and Equipment as well as Stationary Fire Extinguishing Systems in the Transition Process from AFFF Foams to Fluorine-free Foams with Cornelsen's PerfluorAd Technology is executed in 3 Steps 2021): 1. Complete and careful emptying of all System Components (possibly even with partial replacement of Components): Pipes, Hose Lines, Seals, Valves, Pumps, Fittings, Tanks including Partitions and hidden Areas, ... 2. Performing a Flushing of all individual Pipelines with a PerfluorAd Dilution. The last Flushing is carried out with Fresh Water. The visually recognizable Foam Formation serves as an Indicator for the Degree of Cleaning. 3. Treatment of the collected Rinse Water directly on site with a further PerfluorAd application. Off-site Disposal of Rinse Water does not take Place. The PFAS Content of the Rinse Water can already be significantly reduced when using the PerfluorAd Technology exclusively. The steps are identical for the cleaning of equipment of fire brigades and for stationary fire extinguishing systems. In Figure 18 the three individual steps of the cleaning procedure are shown schematically. The cleaning of stationary equipment is shown at the left and the cleaning of fire brigade machines is shown at the right (no illustration available for the first step). 271 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 20: Schematic overview on the three individual steps of the cleaning procedure are shown schematically The cleaning of stationary equipment is shown at the left and the cleaning of fire brigade machines is shown at the right (no illustration available for the first step) 4.13.2. Remaining PFAS concentrations According to several factors: the achievable PFAS residual concentrations in the system depend on Degree of emptying of the entire system (do PFAS deposits still remain in the system after emptying has been completed?) Materials present in the system (plastic, GFK, rubber, ...) Are all components accessible for mechanical cleaning (steam jet, brush, ...) or can adhesions remain in places that cannot be seen? What is the effort involved in replacing "critical components" (e.g. are all seals and plastic parts replaced before cleaning?) If complete emptying is possible and subsequent "bleeding" of PFAS from individual components is impossible, and at least 3 (better 5) rinses with a PerfluorAd dilution 272 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS and a final rinse with fresh water are performed (Depending on the boundary conditions described before, flushing water volumes of >15 to <30 m/vehicle are often required in practice), Considering all of these factors, using the PerfluorAd technology final residual concentration (measured in the final rinse with fresh water) of less than 1.0 g/l total PFAS, very often less than 0.3 g/l to 0.0 g/l can be achieved. Figure 21: Results of the Cleaning of Fire Trucks using PerfluorAd 4.13.3. Costs According to several parameters flow into the pricing, such as size, type, age ... of the vehicle, disposal of the AFFF concentrate to our extent (y/N?), place of cleaning (in our approved installation in Essen or at the customer's?), etc. Depending on these boundary conditions, the costs are usually between 20,000-25,000 Euro/vehicle. These figures include the treatment of all rinsing water and all disposal costs Interview 2021). 4.13.4. Additional information and available case studies A typical PerfluorAd application is the cleaning of fire brigade trucks. For this GmbH is accredited by a German environmental authority. The process takes approx. one working week (Monday to Friday). However, a longer time is needed if components need to be replaced. GmbH is providing the plant and needed personnel. According to MC, several parameters flow into the pricing, such as size, type, age ... of the vehicle, disposal of the AFFF concentrate to our extent (Y/N?), place of cleaning (in our approved installation in Essen or at the customer's?), etc. Depending on these boundary conditions, the costs are usually between 20,000-25,000 Euro/vehicle. These figures include the treatment of all rinsing water and all disposal costs. In the last 6-12 months GmbH is receiving more inquiries from companies with large fire extinguishing systems. MC mentioned that the costs for systems are very hard to extrapolate, as they differ based on the needs of the company. In large fire extinguishing systems, the PFAS-concentrates normally are contained in the tank, tank to mixer pipes, and finally the mixer. Based on the knowledge of MC, PFAS-contamination is not to be expected after the mixer. However, if there have been incidents/training with PFAS-foams then also these systems are expected to be contaminated. In MC's understanding, cleaning of all systems is not required, thus, only PFAS-containing systems (as described before) would need to be cleaned. Based on MCs opinion, a replacement of critical components (plastics, etc.) is advisable to remove PFAS-substances and to prevent future "bleeding" of PFAS-substances. 273 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 1. Complete draining of the foam concentrate (dispose of foam concentrate) 2. Remove foam concentrate residues mechanically and by rinsing with hot (50-60 C) water. All pipes and fittings carrying foam concentrate must also be rinsed during this process. The rinsing process is sufficient when the draining water no longer foams. The flushing water must be disposed of240. 3. The tank, the lines and fittings carrying the foaming agent must be completely filled with water that is as hot as possible. The water must remain in the tank for at least 24 hours. After that, the water must be completely drained and disposed of. 4. The tank, the pipes carrying the foaming agent and the fittings must be completely filled with hot water three more times. The water must remain in the tank for at least 24 hours each time. The rinsing water from these rinsing processes can - if careful procedures are followed - be discharged via the sewage system into the sewage treatment plant. After this cleaning procedure, the tank can be filled with fluorosurfactant-free foam concentrate. The complex cleaning process ensures that the new, fluorosurfactant-free foam concentrate is not contaminated with fluorosurfactants from the tank wall. 4.14.2. Remaining PFAS concentrations According to stakeholder knowledge, the effectiveness of the cleaning is monitored by measurements. A foam concentrate tank can then be released for further use if sufficient cleaning success is guaranteed. As a rule, concentrations below 10 ng/l of each of the 13 standard PFAS241 can be achieved with the cleaning procedure described in the guide and, if necessary, replacement of all accessible seals (LfU-Gierig-Interview 2021). The stakeholder reported, that usually < 10 ng/l, i.e. 10 ppt, related to foam concentrate tanks in fire engines are achievable. LfU does not have any figures for stationary extinguishing systems. LfU also sometimes accepts cleaning efficiencies the range of 100 ng, when special circumstances are to be considered (PFAS-emitting gaskets cannot be replaced). 4.14.3. Costs Costs are available for tank fire engines. Costs are approx.- 100.000-200.000 per engine, when a permanently installed foam concentrate tank is cleaned before refilling with fluorosurfactant-free foam concentrate. 240 Referring to the Lfu-guideline foaming agents containing fluorine surfactants must be disposed of by suitable disposal companies (German waste code number usually 16 10 01* = aqueous liquid waste containing hazardous substances). Certified disposal companies can be researched at www.lfu.bayern.de/abfall/entsorgerfachbetriebe/recherche/index.htm . 241 Measurements based on DIN 38414-14 the German standard methods for the examination of water, waste water and sludge - Sludge and sediments (group S) - Part 14: Determination of selected polyfluorinated compounds (PFC) in sludge, compost and soil - Method using high performance liquid chromatography and mass spectrometric detection 275 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4.14.4. Additional information and available case studies According to stakeholder knowledge, the Munich Fire Department has cleaned its permanently installed foam concentrate tanks according to this guideline. Likewise, other fire departments in Bavaria are likely to have successfully cleaned their foam concentrate tanks according to this procedure (LfU-Gierig-Interview 2021). According to stakeholder knowledge, In Bavaria, there are approximately 1,000 fire engines with permanently installed foam concentrate tanks. Most of these tanks can be sufficiently cleaned with the recommended flushing procedure. The residual quantities are tolerable, especially since a large number of fire extinguisher fills containing PFAS are still in circulation. It would be completely uneconomical to completely replace the vehicle tanks as long as significantly higher inputs of PFAS are present in other areas (LfU-Gierig-Interview 2021). 4.15. Cleaning protocol by Fire Rescue Victoria (FRV) Appliance PFAS Decontamination Project Fire Rescue Victoria and the United Firefighters Union developed a decontaminate procedure for appliances (fire trucks). According to their own knowledge, FRV are the only firefighting authority in the world that has achieved this. Due to the verified success of this PFAS decontamination work, FRV have assisted many other emergency service agencies, to either advise or provide similar decontamination processes and applied safe threshold limits for their respective firefighting appliances. FRV are considered national leaders in the successful implementation of measurable PFAS mitigation work. Background The MFB previously used PFAS containing Aqueous film forming foams (AFFF) as firefighting foam, based on 3M's earlier recommendation of it being safe. Global use of PFAS containing AFFF firefighting foam is the source of approximately 30% of the worlds' PFAS contamination. In 2007, MFB made a decision to replace existing firefighting foam with fluorine-free firefighting foam. This decision was on the basis of concerns relating to firefighter health and environmental issues. MFB then phased out the use of persistent PFAS-containing firefighting foams across its operations. MFB engaged expert independent environmental consultants to analyse PFAS exposure pathways for MFB firefighters on the job. This report was used to inform and develop MFB PFAS threshold limits and prioritise PFAS mitigation work. The MFB (FRV) developed and formally endorsed an `Operational Use of Firefighting Foam Policy' and the use of fluorine free foam. Victorian Environmental Protection Authority (Vic EPA), and Victorian WorkSafe played a part to formalise this policy. By 2014, all MFB firefighting appliances had been converted to only carry fluorine free B Class foam in their foam tanks. Following the MFB establishment of the safe PFAS thresholds, in 2016, MFB embarked on a process to test and decontaminate the MFB firefighting fleet. Further work is currently being done to PFAS decontaminate more recently introduced FRV appliances and monitoring the previous PFAS decontamination work on the earlier MFB appliances. Cleaning Procedure FRV used independent environmental consultants and our industrial cleaning partners, to develop a 32-stage decontamination and verification process targeted at ensuring that the 276 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS appliances, after decontamination, can be safely returned to service. A detailed description of all steps is shown in the annex (see chapter 7.1). In the following each of the steps will be shortly introduced (information taken from a presentation submitted by a stakeholder (FireRescue-Victoria 2021)): 1. Suitable Facility for the PFAS Decontamination process: Fire trucks are taken to a specially constructed decontamination facility, where the removable components (hoses, connectors, ladders etc.) are stripped off for separate decontamination. The trucks are then put into a bund system, where the raw foam is carefully pumped out and the tanks prepared to be flushed and cleaned. 2. Flushing of the tanks: The tanks are carefully flushed with slowly introduced, temperature controlled, water to maximise raw-product foam removal whilst minimising foam creation. Wastewater is collected for future processing and disposal. After removal of the majority of foam product, agitation is introduced to break down and dissolve solidified foam product. Separate, colour-coded pumps and pipelineslines are maintained to ensure that cross-contamination is avoided. Filters, strainers and breathers are carefully dismantled to allow removal of solidified foam product, found to have built-up inside on-board components, wherever there are gaps, welds, connectors, or in joints and gaskets. 3. Cleaning of truck internals: The on-board water pumps are fed by, and feed, an intricate series of pipes, lines and injectors. Cross contamination has been found to be common, and the pumps and feed lines internal to the truck need to be cleaned. 4. Cleaning of delivery systems: Delivery of foam/water mix can be through on-board hose reels, direct to hose systems from the main delivery panels on the side of the trucks, or from what is termed `the monitor', a roof-mounted delivery system. Each of these also needs to be decontaminated. 5. Purging of truck internal lines: A specially designed multi-part manifold is connected to the truck and the internal pump systems. Lines and foam injectors are purged. 6. Cleaning of Onboard components: Truck-mounted hose reels and monitor are decontaminated by flushing with clean water. Ground-spray systems are also flushed. Detachable components are decontaminated separately. Finally, the whole appliance is pressure washed. The interior voids on the truck, where the tanks and pumps sit, is also pressure washed. All washings are collected using a wet-vac system, for subsequent treatment. 7. Cleaning of Removable components: Each truck also has a series of removable components such as fire-fighting hoses, connectors, uptake and transfer hoses that need to be decontaminated. 8. Hose Decontamination: Fire-fighting hoses are decontaminated both externally and internally using a series of specially design hose-washing units. Several lengths of hose are connected to a high-pressure water recycling unit for internal decontamination. This device has a 5,000 litre water tank and a pump capable of high pressure delivery According to FRV, the key to successful decontamination is correct sequencing of operations, and detailed recording of each stage of operations. Each truck decontamination can create between 6,000 - 8,000 liters of wastewater. Wastewater is re-concentrated by passing through a series of activated carbon filters (GAC). It has been possible to strip out the PFAS foam from the waste and achieve sub part-per billion results in the treated wastewater, enabling this to be disposed to trade waste, with the carbon sent for high temperature destruction. 277 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 22: The Decontamination Process - in pictures (taken from (Fire-Rescue-Victoria 2021)) 4.15.1. Remaining PFAS concentrations Figure 23: Achieved PFAS levels after decontamination according to the protocol by FVR, before (blue) and after (red) (taken from (Fire-Rescue-Victoria 2021)) For the remaining PFAS levels, a two-tier decontamination has been chosen. FRV firefighting urban operations and routine training are two very different scenarios. In urban operations, a firefighter might deploy foam in one fire in every 20-50. Water is deployed into a highly modified urban environment. In routine training the same location (a fire training facility) is 278 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS of the tank and concentrate sections of pipework (including proportioners) until no frothing is visible (FPA-AUS 2020). It also requires collection, remediation and safe disposal of all effluent from this washing process. Cleaning Procedure FPA Australia recommends the following process when cleaning foam tanks or changing out existing C8 foams: 1. Decant existing C8 foam into suitable storage containers, which are also bunded and clearly marked for incineration/destruction. 2. Thoroughly flush system with water and collect effluent in suitable storage containers/tankers, identifying contents. The use of hot water may facilitate cleaning. 3. Using suitable remediation technologies, flushed foam solution and effluent should be treated to concentrate the PFAS into as small a volume as practical and should be held separately and labelled prior to disposal/destruction. 4. Analyse clean water for residual PFAS levels, before any release for re-use to the sewer/environment to ensure local regulatory requirements are met. This is likely to require temporary storage in large clean tanks without any previous PFAS usage or potential pre-existing PFAS contamination. 5. Send concentrated PFAS containing materials for disposal/destruction in accordance with local regulatory requirements. 4.16.2. Remaining PFAS concentrations To avoid the possibility of contamination, the tank should not be filled with the replacement foam until the results of this testing are available and confirm sufficiently low levels acceptable to the local environmental regulator. It is recommended a sample of the "clean" effluent be tested by a NATA accredited laboratory for traces of PFOS/PFHxS/PFOA to determine a baseline level of contamination for future reference and to confirm the storage is essentially "PFOS, PFHxS and PFOA free" down to the levels specified in the Queensland Policy. 4.16.3. Costs There is no information available on costs for this procedure. 4.16.4. Additional information and available case studies There is no information available on case studies. 4.17. Cleaning protocol by the Australian DoD 4.17.1. Background The Aircraft Rescue & Firefighting (ARFF) foam transition project will transition of all Army, Air Force and Broad spectrum firefighting vehicles to a suitable Fluorine Free Foam (F3) product (DoD-AUS 2020b). The ARFF Vehicle Foam Transition Sub Working Group will manage this project under delegation from the One Defence Firefighting Foam Management Working Group. As seen in the figure (Figure 22) below, the cleaning procedure relies on the set up of 10 cleaning hubs, where over 100 vehicles will be cleaned. The procedure and the description are very detailed and equipped with further documentation by photos. In the following only a brief overview will be given. 280 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 24: Overview on the Aircraft Rescue & Firefighting (ARFF) foam transition project (DoD-AUS 2020a) Cleaning Procedure FPA Australia recommends the following process when cleaning foam tanks or changing out existing C8 foams: 1. Decanting Aqueous Film Forming Foam (AFFF). ARFF vehicles will be decanted of AFFF prior to chain of custody being taken by the Hub Supervisor, and cleaning activities comm Continuous flush. 2. Sample baseline 3. Vehicle CES soaking 4. Outlet and hose flushing 5. Sample for validation 6. Re-fill with F3. ARFF vehicles will be re-filled with F3 upon completion of the required cleaning activities set out in step 4. The Hub Supervisor will apply a colour coded zip tie to cleaned CES items associated with the vehicle to identify them as F3 only. 7. Proportioner Calibration and Return to Service Testing (RTS): The User Units will conduct vehicle foam performance tests (including proportioner calibration) inside, or at, the designated foam test facility or area as per existing testing arrangements. 8. Restart next vehicle 9. Vehicle validation against cleaning criteria There is a checklist, which needs to be checked before the vehicle can leave the hub. 281 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 25: Cleaning procedure phases in accordance to the Queensland DoD (DoD-AUS 2020c) 4.17.2. Remaining PFAS concentrations There is no information available for the remaining PFAS concentration. 4.17.3. Costs There is no information available on costs for this procedure. 4.17.4. Additional information and available case studies There is no information available on case studies. 4.18. Cleaning protocol by Deutschland (WFVD) Werkfeuerwehrverband According to WFVDs homepage242, the plant and company fire departments as well as the company fire protection officers have a special task within the framework of the organization serving fire protection. In addition to the general fire protection tasks, the company fire protection organization must take into account the special company risks. Preventive fire protection, hazard prevention and company rescue services are of particular importance. As a result of the need to adapt to company-specific conditions, the principles and guidelines established for public fire departments cannot be transferred to plant and company fire departments without further ado. The plant and company fire departments as well as the company fire protection officers therefore need their own organization for the purpose of representing their interests and exchanging experience. 242 See here, last accessed 24.03.2021 282 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Now that the relevant organizations have largely been formed at the level of the German states, the "Bundesverband Betrieblicher Brandschutz/Werkfeuerwehrverband Deutschland e.V.", hereinafter referred to as WFVD, is the appropriate representative body. 4.18.1. Background In 2014 it was noticed that PFAS levels in the foam concentrate of two fire apparatus at an industrial facility is higher than expected (for example 30,000g/kg 6:2FTS). The purchased foam concentrate used in this tank was a C6 based AFFF (not based on PFOS, see Figure 3 for a PFAS analysis of the new product) and it did not exceed the threshold of 10,000g/kg PFOS. However, the foam concentrates in the tanks of the of two fire apparatus exceeded PFOS levels. This was traced back to a cross contamination from a PFOS-based foam concentrate (3M Lightwater) that was in the same foam concentrate tanks before. It was supposed that during transition from C8 (PFOS based) foam to C6 (PFHxA precursor based) foam, the tanks were not cleaned sufficiently. Residues of 3M Lightwater contained high amounts of PFOS and contaminated the new PFOS-free AFFF. It was decided to develop a foam concentrate tank cleaning procedure, clean the tanks accordingly and transition to fluorine free foam. According to WFVD, from these cases it can be concluded that the described foam concentrate tank cleaning procedure is an effective method to clean tanks for firefighting foam concentrate when transitioning from PFAS-based to fluorine free foam. It is simple enough to be carried out by fire brigades themselves or for example by companies that specialize in industrial cleaning. It sufficiently reduces PFAS levels below applicable standards and is adjustable in case the results do not meet expectations. Cleaning Procedure The foam concentrate tank cleaning procedure is a relatively simple process that in many cases can be carried out by fire services themselves. Basically, it comprises a series of flushing with water, after the tank is emptied. Main challenge in this process is to avoid spills and contamination of equipment outside the foam concentrate tank. During step 2 and 3 the residues of the foam concentrate will cause foaming inside the tank. The overflow of that foam should be avoided to not cause any contamination outside of the tank. Further attention should be paid to a proper disposal of the old foam concentrate and any rinsing water. The standard disposal method would be high temperature incineration in a facility that is able to handle PFAS waste. WFVD recommends the following foam concentrate tank cleaning procedure: 1. Step 1: a. Empty foam concentrate tank, pump and piping b. Dispose foam agent through high temperature incineration 2. Step 2: a. Fill tank with warm water (60-70C) (half full to avoid overflow of foam) b. Drive with apparatus for a 30minutes to allow contact of water with the whole inner tank surface c. Pump water with foam pump in a loop for about 30 minutes d. Empty tank, pump and piping e. Destroy foam inside tank with water and a very fine nozzle and empty tank again f. Dispose water through high temperature incineration 3. Step 3: a. Repeat step 2 one time 283 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 4. Step 4: a. Fill tank with water b. Pump water with foam pump in a loop for about 30 minutes c. Take a water sample d. Analyse water sample for PFAS e. Repeat Step 3 if results of PFAS analysis are not sufficient f. Dispose water through high temperature incineration 5. Step 5 a. Drain any rinsing water from tank, pump and any pipes b. Dry tank as good as practically possible c. Fill tank with new foam concentrate 4.18.2. Remaining PFAS concentrations According to WFVD, the efficacy of the foam concentrate tank cleaning procedure can be assessed when looking at Figure 24 and Figure 25. The highest remaining PFAS-substances reported in Figure 24 are 6:2 FTS with 0,98 g/L (0.00098 ppm) and PFOS with 0,81 g/L (0.00081 ppm). The highest remaining PFASsubstances reported in Figure 25 are PFOS with 42 g/L (0.042 ppm) and PFOA with 1,2 g/L (0.0012 ppm). If all reported PFAS-substances are added a remaining concentration of 57,5 ppm (0.057 ppm)is measured. These are PFAS analyses of the rinsing water from step 4 of the procedure. As these are analyses for PFAS in water the detection limit for PFAS is lower than in the analyses for PFAS in foam concentrate. The analyses show that cleaning is effective with dilution factors varying between 100 and 100,000. While PFAS can still be detected in the rinsing water in all cases they are lower than current applicable thresholds for PFOS (not further commented by the stakeholder but most likely 10 ppm according to POP-regulation) and PFOA (not further commented by the stakeholder). If the results do not meet expectations step 2/3 can be repeated until levels are sufficiently low. It has to be noted that the water analyzed in step 4 will also be disposed and that PFAS levels can be assumed to be even lower when the new foam concentrate is filled in. Figure 26: PFAS Analysis of rinsing water from apparatus "TMB" from step 4 of tank cleaning procedure (note, that the detection limit is lower as this is an analysis for PFAS in water as opposed to PFAS in foam concentrate in other figures) 284 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS Figure 27: PFAS Analysis of rinsing water from apparatus "PTLF II" from step 4 of tank cleaning procedure (Note 1: The detection limit is lower as this is an analysis for PFAS in water as opposed to PFAS in foam concentrate in other figures Note 2: This apparatus is also referred to as "TroTSLF 2" or "PTLF 2") 4.18.3. Costs Costs for the cleaning of a foam concentrate tank of a fire apparatus highly depend on disposal costs for the foam concentrate and flush water, summing up to 50% of the total costs in this example (4000 ). It is estimated that the volume of the flush water is three times that of the tank volume. Other costs, like work hours are likely to be independent from tank size (unless deviating to a greater extent from this example). In this case study the work was done by the industrial fire brigade itself, so that no external costs arose for work hours. Figure 28: Estimated costs for the cleaning of a 1 m foam concentrate tank with the described cleaning procedure 4.18.4. Additional information and available case studies Three years after the cleaning of the tanks, the foam concentrate was analysed for PFAS again. Except for one PFAS in the apparatus all PFAS are below the detection limit and, most important, below the applicable threshold for PFOS and PFOA. The reason for the 56g/kg of 6:2 FTS in apparatus "TMB" are not known for sure. Possible explanations are: Cross contamination from residues of old foam concentrate, contamination of the sample or measuring error. The procedure is also explained in a video available at Youtube (English and German version). The stakeholder reported there are some mistakes concerning the values in the English version. 285 ANNEX XV RESTRICTION REPORT - PFAS IN FIREFIGHTING FOAMS 286