Document x5nR7dmME92zXwmVvN6nqe7x0

THE COST OF INACTION A socioeconomic analysis of environmental and health impacts linked to exposure to PFAS The cost of inaction A socioeconomic analysis of environmental and health impacts linked to exposure to PFAS Gretta Goldenman, Meena Fernandes, Michael Holland, Tugce Tugran, Amanda Nordin, Cindy Schoumacher and Alicia McNeill TemaNord 2019:516 The cost of inaction A socioeconomic analysis of environmental and health impacts linked to exposure to PFAS Gretta Goldenman, Meena Fernandes, Michael Holland, Tugce Tugran, Amanda Nordin, Cindy Schoumacher and Alicia McNeill ISBN 978-92-893-6064-7 (PRINT) ISBN 978-92-893-6065-4 (PDF) ISBN 978-92-893-6066-1 (EPUB) http://dx.doi.org/10.6027/TN2019-516 TemaNord 2019:516 ISSN 0908-6692 Standard: PDF/UA-1 ISO 14289-1 Nordic Council of Ministers 2019 Cover photo: Scanpix Disclaimer This publication was funded by the Nordic Council of Ministers. However, the content does not necessarily reflect the Nordic Council of Ministers' views, opinions, attitudes or recommendations. Rights and permissions This work is made available under the Creative Commons Attribution 4.0 International license (CC BY 4.0) https://creativecommons.org/licenses/by/4.0 Translations: If you translate this work, please include the following disclaimer: This translation was not produced by the Nordic Council of Ministers and should not be construed as official. The Nordic Council of Ministers cannot be held responsible for the translation or any errors in it. Adaptations: If you adapt this work, please include the following disclaimer along with the attribution: This is an adaptation of an original work by the Nordic Council of Ministers. Responsibility for the views and opinions expressed in the adaptation rests solely with its author(s). The views and opinions in this adaptation have not been approved by the Nordic Council of Ministers. Third-party content: The Nordic Council of Ministers does not necessarily own every single part of this work. The Nordic Council of Ministers cannot, therefore, guarantee that the reuse of third-party content does not infringe the copyright of the third party. If you wish to reuse any third-party content, you bear the risks associated with any such rights violations. You are responsible for determining whether there is a need to obtain permission for the use of third-party content, and if so, for obtaining the relevant permission from the copyright holder. Examples of third-party content may include, but are not limited to, tables, figures or images. Photo rights (further permission required for reuse): Any queries regarding rights and licences should be addressed to: Nordic Council of Ministers/Publication Unit Ved Stranden 18 DK-1o61Copenhagen K Denmark Phone +45 3396 0200 M@norden.org Nordic co-operation Nordic co-operation is one of the world's most extensive forms of regional collaboration, involving Denmark, Finland, Iceland, Norway, Sweden, and the Faroe Islands, Greenland and Aland. Nordic co-operation has firm traditions in politics, economics and culture and plays an important role in European and international forums. The Nordic community strives for a strong Nordic Region in a strong Europe. Nordic co-operation promotes regional interests and values in a global world. The values shared by the Nordic countries help make the region one of the most innovative and competitive in the world. The Nordic Council of Ministers Nordens Hus Ved Stranden i8 DK-1o61Copenhagen K, Denmark Tel.: +45 3396 0200 www.norden.org Download Nordic publications at www.norden.org/nordpub Contents Preface ......................................................................................................................................7 Purpose of this report ..........................................................................................................7 Disposition ..........................................................................................................................7 Scope and limitations ..........................................................................................................7 Financing and workforce..................................................................................................... 8 Acknowledgments.............................................................................................................. 8 Summary................................................................................................................................. 11 Conclusions .......................................................................................................................11 Methodology .....................................................................................................................12 Impact pathways (the case studies)....................................................................................13 Health-related costs to society...........................................................................................14 Non-health (environment-related) costs to society ............................................................15 Abbreviations used ................................................................................................................. 19 1. Introduction.......................................................................................................................23 2. The regulatory framework as baseline in relation to PFAS..................................................27 2.1 Guideline values for protection of health related to PFAS exposure ........................ 28 2.2 Other regulatory actions underway .........................................................................31 3. Methodology to assess environment and health-related costs...........................................33 3.1 Health-related costs................................................................................................35 3.2 Environment-related costs ......................................................................................41 4. The case studies ................................................................................................................51 4.1 Case Study 1: Exposures due to production of PFAS ................................................51 4.2 Case Study 2: Exposures due to manufacture and commercial use of PFAS- containing products ............................................................................................... 58 4.3 Case Study 3: Contamination from use of aqueous film-forming foams.................. 64 4.4 Case Study 4: Exposures during the use phase of PFAS-treated products.................73 4.5 Case Study 5: Impacts at end of life of PFAS-treated products ................................ 82 4.6 Other costs related to PFAS contamination............................................................ 90 4.7 Summary of case study findings ............................................................................. 91 5. Estimates of costs of inaction linked to exposure to PFAS ................................................. 95 5.1 Health-related costs of exposure to PFAS............................................................... 95 5.2 Non-health costs of environmental contamination with PFAS ............................... 107 6. Conclusions .....................................................................................................................125 6.1 Findings and discussion.........................................................................................125 6.2 Next steps and proposals for further studies ..........................................................130 References ............................................................................................................................ 131 Sammanfattning ...................................................................................................................143 Slutsatser ........................................................................................................................143 Metod 144 Spridningsvgar (fallstudier) ............................................................................................145 Hlsorelaterade samhllskostnader .................................................................................146 Icke hlsorelaterade samhllskostnader (miljrelaterade)................................................147 The cost of inaction linked to PFAS exposure 5 Annex 1: National / state screening for PFAS contamination.................................................. 151 Denmark ......................................................................................................................... 151 Sweden........................................................................................................................... 153 Finland............................................................................................................................ 154 Belgium (Flanders) .......................................................................................................... 155 The USA (national monitoring) ........................................................................................ 155 The US state of Michigan ................................................................................................ 156 Annex 2: Health impacts - additional information on evidence and calculations .....................157 Part 1: Estimating health impacts in exposed populations.................................................157 Part 2: Key epidemiological studies ................................................................................. 164 Annex 3: Data used for environment-related cost calculations ............................................... 165 Part 1: The costs gathered via the case studies and additional research............................ 165 Exchange rates used........................................................................................................ 165 Monitoring to assess PFAS contamination where it is suspected...................................... 166 Provision of a temporary uncontaminated drinking water supply..................................... 168 Upgrading of water treatment works and ongoing costs for maintenance and replacement and disposal of filters ............................................................................................ 169 Excavation and treatment of soils.....................................................................................173 Health assessments where contamination is found.......................................................... 174 Part 2: Data used in the aggregation of costs ................................................................... 176 Part 3: Full cost estimates by country............................................................................... 183 6 The cost of inaction linked to PFAS exposure Preface Purpose of this report There is a growing demand for monetary values within chemical policy. The purpose of this report is to estimate the costs for society related to the negative effects and impacts on human health and environment due to the exposure to PFAS and C4-14 non-polymer fluoro-surfactants in particular. The purpose is also to highlight the economic case for taking effective and timely action to manage the risks of negative impacts. The information in this report is intended to be used to raise awareness on the costs and long-term problems that the use of PFAS may cause for the environment and human health. The use of monetary values provides an additional important basis for strategical decisions within chemical agencies both at the national level as well as on the EEA level. Disposition The report is divided into three main parts. The first part provides a regulatory baseline and outlines the methodology to assess the socioeconomic costs related to the negative impacts on the environment and human health. The second part presents five case studies chosen for this study. They are aimed at illustrating the key pathways for impacts from PFAS and to gather information on actual costs incurred by society in reducing exposure to PFAS. The third part presents the estimates of health and environmentrelated costs of inaction linked to exposure to PFAS as well as the aggregated costs of inaction. Scope and limitations This study focuses on the C4-14 non-polymer fluoro-surfactants with the aim of providing a monetised estimate of total damage to health and the environment associated with PFAS exposures in the European Economic Area (EEA). The report therefore focuses on costs of inaction in the EEA countries. It uses data specific to Nordic countries when available, but also draws cost data from other European countries, the USA and Australia, where relevant. Not all costs can be quantified and monetised; some costs are therefore assessed qualitatively. The study considers only the socioeconomic costs incurred by society due to impacts from PFAS exposures. It does not include or monetise costs for business such as for example substitution costs. The cost of inaction linked to PFAS exposure 7 Financing and workforce The study was carried out by Milieu Consulting in Brussels. Authors include Gretta Goldenman, Meena Fernandes, Tugce Tugran, Amanda Nordin, Cindy Schoumacher and Alicia McNeill from Milieu and Michael Holland from EMRC. The health economic models and calculations were developed and described by Meena Fernandes. The methodological framework for estimating costs for society, as well as the environmental economic models and calculations, were developed and described by Michael Holland. The Nordic Chemical Group (NKG) has been the project principal. The steering group under NKG comprised members from Sweden, Denmark, Norway and Iceland, including Toke Winther and Lars Fock (The Danish Environmental Protection Agency), Audun Heggelund (Norwegian Environment Agency), Signe Krarup (The Danish Ministry of Environment and Food) and sa Thors (The Swedish Chemicals Agency). Experts supporting the steering group for parts of the project included Jenny Jans, Inger Cederberg, Mattias Carlsson Feng, Daniel Borg (The Swedish Chemicals Agency) and sak Sigurjn Bragason (The Environment Agency of Iceland). Acknowledgments The authors of the study wish to thank the members of the Steering Group for their contributions and support for this project. The authors also acknowledge and thank the following persons for the information and guidance they provided at key moments in the course of the study. Responsibility for the analysis, interpretation of data and conclusions drawn rests solely with the authors. Background and case study information Julie Ng-A-Ham & Korienke Smit (Netherlands Institute for Public Health and the Environment); Greet Schoeters (Flanders Institute of Technology); Dr Ian Cousins (Stockholm University), Dr Martin Scheringer (Swiss Federal Institute of Technology); Dr Zhanyun Wang (Swiss Federal Institute of Technology); Dr Xenia Trier (European Environment Agency); Bastian Zeiger (European Pollutant Release and Transfer Register); Erwin Annys (CEFIC); Patricia Vangheluwe (Plastics Europe); Amy Peterson (Department of Environmental Quality, State of Michigan); Tom Bruton (Green Science Policy Institute); Roland Weber (POPs Environmental Consulting); Lena Vierke (German Umweltbundesamt). 8 The cost of inaction linked to PFAS exposure Costs of health impacts Dr Tony Fletcher (London School of Hygiene and Tropical Medicine); Dr Kristina Jakobsson (Lund University); Dr Leo Trasande (New York University); Dr Jamie C. Dewitt (East Carolina University); Philippe Grandjean (Harvard School of Public Health/ University of Southern Denmark); Emma Halldin Ankarberg (Swedish National Food Administration); Marina Mastrantonio (Italian National Agency for New Technologies, Energy and Sustainable Economic Development); Antonia Calafat (Centers for Disease Control and Prevention) Costs of non-health (environmental) impacts Dr Detlef Knappe (North Carolina University); Dr David Sunding (University of California, Berkeley); Reiner Shlmann (Landratsamt Rastatt Umweltamt); Dr Frans Lange (DVGW); Sara von Ehrenheim, Gran Carlsson & Johan Eriksson (Uppsala municipality); Karl Lilja (Swedish EPA); Olaf Kaspryk & Lorena Rodriguez (Stadtwerke Rastatt GmbH); Lena Wennberg (Swedavia); Gril Aasen Slinde (Norwegian Geotechnical Institute); Dorte Herzke (Norwegian Institute for Air Research); Ida Schyberg (Ronneby Milj & Teknik AB); Magnus Olofsson (Ronneby Municipality); Michael Hopgood & Niklas Lwegren (Swedish Transport Administration); Jakob Gille (Swedish Armed Forces) (Sarah Bull (WrC); Griet Van Gestel (Flanders Waste Management Agency); Dr. Karlin Stark (Regierungsprsidium Stuttgart Landesgesundheitsamt). The cost of inaction linked to PFAS exposure 9 10 The cost of inaction linked to PFAS exposure Summary This study investigates the socioeconomic costs that may result from impacts on human health and the environment from the use of PFAS (per and polyfluoroalkyl substances). Better awareness of the costs and long-term problems associated with PFAS exposure will assist authorities, policy-makers and the general public to consider more effective and efficient risk management. The production of PFAS, manufacture and use of PFAS-containing products, and end-of-life disposal of PFAS have resulted in widespread environmental contamination and human exposure. PFAS have been found in the environment all around the world and almost everyone living in a developed country has one or more PFAS in his/her body. Because of the extreme persistence of PFAS in the environment, this contamination will remain on the planet for hundreds if not thousands of years. Human and environmental exposure will continue, and efforts to mitigate this exposure will lead to significant socioeconomic costs - costs largely shouldered by public authorities and ultimately taxpayers. The focus of this study is on the costs of inaction with respect to regulation of PFAS in the countries comprising the European Economic Area (EEA). Costs of inaction are defined as the costs that society will have to pay in the future if action is not taken to limit emissions of PFAS today. The PFAS covered in this study are the C4-14 non-polymer fluorosurfactants. The goal for the study has been two-fold: 1. to establish a framework for estimating costs for society related to negative impacts on health and the environment associated with PFAS exposure; and 2. to provide monetary values for those societal costs, documented by case studies. Conclusions The work of estimating the health and environment-related costs to society related to PFAS exposure has relied on the development of assumption-based scenarios. This reflects the limited data available in the academic literature, government documents and press reports. Whilst the uncertainties of the analysis need to be acknowledged, it is also important to recognise that, for several issues, there is little or no uncertainty: The cost of inaction linked to PFAS exposure 11 1. PFAS are ubiquitous in the environment, and almost all people have PFAS in their bodies today. Monitoring in both Sweden and the USA concludes that around 3% of the population are currently exposed above proposed limit values, primarily through contamination of drinking water but also via other sources; 2. Many sources of PFAS exposure exist, linked to specialist applications (e.g. AFFFs for firefighting at airports and some industrial locations) and non-specialist uses (e.g. use in consumer goods such as pizza boxes, clothing and cosmetics); 3. Non-fluorinated alternatives for many of these uses are already on the market, and therefore certain uses of PFAS can be reduced; 4. The costs for remediating some cases of contamination run to many millions of EUR. Total costs at the European level are expected to be in the hundreds of millions of EUR as a minimum; 5. A large and growing number of health effects have been linked to PFAS exposure and evidence is mounting that effects occur even at background level exposures. Current and proposed limit values for drinking water may be further reduced in recognition of growing information on, health and environmental risks. This would increase the costs of environmental remediation estimated here. As explained throughout the study, the calculations rest on a number of assumptions, though these have been checked against e.g. data on costs incurred to ensure that they are linked to real-world experience. As more information becomes available, calculations will become more precise. Moreover, these findings are conservative. The figures are likely to get larger, in that the numbers of PFAS on the market and the volumes produced keep increasing. Further inaction will lead to more sources of contamination, more people exposed, and higher costs for remediation. The longer that PFAS contamination remains in the environment without remediation, the wider it will spread and the greater the quantity of soil or groundwater that will need to be decontaminated. Methodology Two methodologies have been developed, one for estimating health-related costs, the other for estimating costs of environmental remediation. Both methodologies are based on cases concerning exposure to PFAS. Data from the Nordic countries have been used when available, but the estimates also draw on cost data from other European countries, the USA and Australia, where relevant. 12 The cost of inaction linked to PFAS exposure Impact pathways (the case studies) Five case studies following the life-cycle of PFAS, from their production and use in product manufacturing, to the product's use and end-of-life disposal are used to illustrate how exposures to humans and the environment occur. Other instances of PFAS contamination provide additional data on direct costs incurred. Case Study 1 considers exposures due to the production of PFAS in Europe. It reviews pollution linked to the Chemour factories in Dordrecht, Netherlands, the Miteni facility in the Veneto region of Italy, and the 3M plant near Antwerp, Belgium. The study estimates that up to 20 facilities actively produce fluorochemicals in Europe, that these facilities are significant sources of PFAS released to the environment, and that the exposure of workers at these plants is high. The impacts from the manufacture and commercial use of PFAS-containing products are the focus of Case Study 2. Industrial activities with the potential to release PFAS to the environment include textile and leather manufacturing; metal plating, including chromium plating; paper and paper product manufacturing; paints and varnishes; cleaning products; plastics, resins and rubbers; and car wash establishments. The study assumes that a range of 3% to 10% of these facilities use PFAS. The study did not identify any fluorochemical production facilities in the Nordic countries. However, Eurostat statistics indicate that other industrial activities with the potential to release PFAS to the environment do take place in the region, such as metal plating and manufacture of paper products. Case Studies 3 and 4 consider the use phase of PFAS-containing products. Case Study 3 examines exposure to PFAS-containing aqueous film-forming foams (AFFFs) used in firefighting drills and to extinguish petroleum-based fires. The AFFFs have contributed to groundwater contamination, especially around airports and military bases. Nearby communities have been affected by elevated levels of PFAS in their drinking water. Case Study 4 looks at PFAS-treated carpets, PFAS-treated food contact materials, and cosmetics as examples of how a product's use is likely to lead to direct human exposure through ingestion and dermal absorption. The use of products also result in releases to the environment when the product is washed off or laundered, entering sewers and treatment plants, and eventually waterways. Case Study 5 looks at end-of-life impacts of PFAS-treated products. Municipal waste incineration may destroy PFAS in products if 1000 C operating temperatures are reached. If landfilled, the PFAS will remain even after the product's core materials break down. The compounds will eventually migrate into liquids in the landfill, then into leachate collection systems or directly into the natural environment. They may then contaminate drinking water supplies, be taken up by edible plants and bioaccumulate in the food chain. The cost of inaction linked to PFAS exposure 13 Health-related costs to society To calculate health-related costs to society, the researchers looked for consensus regarding health endpoints affected by exposure to PFAS. Reviews of the scientific evidence have reached contradictory conclusions about the relevant health endpoints of human exposure to PFAS. However, some consensus has emerged concerning liver damage, increased serum cholesterol levels (related to hypertension), decreased immune response (higher risk of infection), increased risk of thyroid disease, decreased fertility, pregnancy-induced hypertension, pre-eclampsia, lower birth weight, and testicular and kidney cancer. The methodology draws upon risk relationships developed in the course of specific epidemiological studies for populations exposed to PFAS at different levels. Workers exposed to PFAS in the workplace were used to exemplify a high level of exposure. Communities affected by PFAS, e.g. because of proximity to manufacturing sites or sites where fluorinated AFFFs were used, were assumed to have been exposed at a medium level; this level of exposure was assumed to have been experienced by 3% of the European population. The general population was considered to have experienced exposure at low (background) levels. Table 1 provides an overview of the estimated annual costs for just a few health endpoints where risk ratios were available for affected populations. For example, the annual health-related costs for the elevated risk of kidney cancer due to occupational exposure to PFAS was estimated to be on the order of EUR 12.7 to EUR 41.4 million in the EEA countries. The estimated costs were substantially higher for elevated and background levels of exposure due to the greater number of persons affected. The total annual health-related costs, for the three different levels of exposure, was found to be at least EUR 2.8 to EUR 4.6 billion in the Nordic countries and EUR 52 to EUR 84 billion in the EEA countries.1 Despite the high level of uncertainty and the assumptions underlying the calculations, the findings suggest that the health-related costs of exposure to PFAS are substantial. 1 The health-related costs due to occupational exposure to PFAS in the Nordic countries was not estimated due to an absence of information about the number and location of chemical production plants or manufacturing sites. 14 The cost of inaction linked to PFAS exposure Table 1: Estimates of annual health impact-related costs (of exposure to PFAS) Exposure level "Exposed" population and source Health endpoint Occupational (high) Elevated (medium) Workers at chemical production plants or manufacturing sites Communities near chemical plants, etc. with PFAS in drinking water Kidney cancer All-cause mortality Low birth weight Infection Background (low) Totals Adults in general population (exposed via consumer products, background levels) Hypertension Nordic countries All EEA countries Population at risk n.a. Annual costs n.a. Population at risk 84,000- 273,000 Annual costs EUR 12.7-41.4 million 621,000 8,843 births 45,000 children 10.3 million EUR 2.1- 2.4 billion 136 births of low weight 84,000 additional days of fever EUR 0.7- 2.2 billion 12.5 million 156,344 births 785,000 children 207.8 million EUR 41-49 billion 3,354 births of low weight 1,500,000 additional days of fever EUR 10.7-35 billion Nordic countries EUR 2.8-4.6 billion All EEA countries EUR 52-84 billion Some overlap occurs in the figures above, because workers and affected communities are also exposed to background levels of PFAS. At the same time, these costs are likely to be underestimates due to the lack of epidemiological-based risk relationships for calculating other health endpoints and related costs. Non-health (environment-related) costs to society The second methodology compiled information on direct costs incurred by communities taking measures to reduce PFAS exposure through remediation of drinking water. Based on these direct costs, ranges of costs per persons affected or per case were developed. These unit costs then became the foundation for aggregating the costs of remediation when environmental contamination, e.g., PFAS concentrations in drinking water, reach certain levels. It should be noted that the ranges are broad, even when normalized against population. The approach to derive ranges for the mean is dependent on the amount of data available. For the costs of water treatment, for example, several estimates were available, and in such cases it is unlikely that the true mean will be at either extreme of the range from the studies. Therefore, it is reasonable to truncate the observed range, for example by removing estimates that are sufficiently removed from other data as to be considered outliers. For some costs, however, very few estimates are available, each of which may be equally valid for representation of the average: in such a case the observed range in values is adopted as the range of plausible mean values. Where no range is available from the studied literature, a range has been estimated. For example, the range of +/-90% is used for establishing a health assessment regime (here considered as a non-health cost as it deals with management of the problem, rather The cost of inaction linked to PFAS exposure 15 than impacts on the health of society). In this example, the range is extremely broad for two reasons, first because of the lack of data available and second because of the potential for variation in the implementation of a health assessment programme. As with the health-based estimates, the study assumes that 3% of the European population is exposed to drinking water with PFAS concentrations over regulatory action levels, such that the water treatment works serving them will require upgrading and maintenance over the next 20 years. The assumption of 20 years reflects potential for remediation to resolve problems perhaps through decontamination or the use of alternative supplies, or the potential for remedial action to persist for many years. Recognising the uncertainties that exist in the analysis and the available data, costs of remediation have been quantified using a scenario-based approach. For each scenario a number of parameters are specified, relating for example to the size of the affected population and the duration of maintenance works. Table 2 shows the range of costs for the various categories of actions related to environmental remediation. Table 2: Summary of estimates of mean cost data for non-health expenditures, 20 years Action taken when PFAS Unit found Best estimate Range from studies Adopted range Monitoring - checks for contamination due to industrial or AFFF use Health assessment (including biomonitoring) Provision of temporary uncontaminated supply Provision of a new pipeline Upgrading water treatment works (capital) Upgrading water treatment works (maintenance) Excavation and treatment of soils - contamination from industrial or AFFF use Cost per water sample tested Cost/case of contamination Cost/person Total biomonitoring and health assessment per case where considered appropriate Cost/person Cost/person Cost/person Cost/person Cost/kg PFAS Cost/case EUR 340 EUR 50,000 EUR 50 EUR 3.4 million No relevant data EUR 800 EUR 300 EUR 19 EUR 280,000 EUR 5 million EUR 278-402 EUR 5,200-5.8 million No range EUR 2.5 million- 4.3 million EUR 37-5,000 EUR 8-2,200 EUR 8-30 EUR 100,000- 4.3 million EUR 100,000-3 billion EUR 278-402 EUR 25,000- 500,000 EUR 5-95 (+/-90%) EUR 1 million-5 million EUR 100-1,500 EUR 18-600 EUR 8-30 EUR 100,000-1 million EUR 300,000-50 million 16 The cost of inaction linked to PFAS exposure In Table 3 the range of costs for the various categories of actions related to environmental remediation for the five Nordic countries are shown. The overall range of costs is EUR 46 million - 11 billion. Table 3: Detailed breakdown of ranges for non-health costs to the Nordic countries, assuming that 1 to 5% (best estimate 3%) of the population is exposed above a statutory limit and that water treatment is required over a 20 year period N people affected (3%) Screening and Health as- monitoring sessment Upgrade treatment works and maintenance Soil remediation Total Denmark 170,000 Finland 160,000 Iceland 10,000 Norway 160,000 Sweden 290,000 Nordic total 790,000 EUR 70,000- 8.3 million EUR 250,000- 22 million EUR 10,000- 900,000 EUR 170,000- 20 million EUR 480,000- 47 million EUR 280,000-27 million EUR 270,000-26 million EUR 20,000-1.6 million EUR 260,000-25 million EUR 490,000-46 million EUR 7.4 million-274 million EUR 0-798 million EUR 8 million- 1.1 billion EUR 7.2 million-265 million EUR 2.2 million- 2.1 billion EUR 10 million- 2.4 billion EUR 400,000-1.6 million EUR 100,000- 86 million EUR 1 million- 105 million EUR 6.8 million-250 EUR 1.6 million- EUR 9 million- million 1.9 billion 2.2 billion EUR 13 million-472 million EUR 4.3 million- EUR 18 million- 4.5 billion 5.1 billion EUR 46 million- 11 billion The cost estimates provided in the table are likely to be more robust at the aggregate, European level than at the national level. Table 4 provides aggregated costs covering environmental screening, monitoring (where contamination is found), water treatment, soil remediation and health assessment for the five Nordic countries and for the other EEA countries and Switzerland. Table 4: Aggregated costs covering environmental screening, monitoring where contamination is found, water treatment, soil remediation and health assessment Best estimate Low High Denmark Finland Iceland Norway Sweden Other EEA+CH Total EUR 145 million EUR 214 million EUR 12 million EUR 194 million EUR 423 million EUR 15,9 billion EUR 16.9 billion EUR 8 million EUR 10 million EUR 1 million EUR 9 million EUR 18 million EUR 776 million EUR 821 million EUR 1.1 billion EUR 2.4 billion EUR 105 million EUR 2.2 billion EUR 5.1 billion EUR 159.9 billion EUR 170.8 billion Parallel calculations for all 31 EEA Member Countries and Switzerland arrive at a range of costs for environmental remediation totalling EUR 821 million to EUR 170 billion. The The cost of inaction linked to PFAS exposure 17 lower and upper bounds should be considered illustrative because of the limited information available. However, based on the literature review, there is a firm basis for concluding that the lower bound estimates would be exceeded. A best estimate in the order of EUR 10-20 billion is certainly plausible. The potential for higher costs is also possible: An estimate of the costs for one case identified in the course of the research, concerning the town of Rastatt in Baden-Wurttemberg in Germany is in the range of EUR 1 to 3 billion, with the estimated extent of the problem being seen to increase over time. The source of contamination in this case is understood to be contaminated waste paper materials that were spread on agricultural land, demonstrating that serious problems are not always linked to airfields and PFAS manufacture. A number of other costs related to PFAS contamination are outside the scope of the quantification carried out in this report. These include loss of property value, reputational damage to a polluting company, ecological damage and the costs incurred by public authorities in responding to affected communities - including public outreach, surveys of contamination and remedial measures. 18 The cost of inaction linked to PFAS exposure Abbreviations used 6:2 FTS AFFF ATSDR BB/CC CA DTSC CAS CLH CLP CMR C8 D4/D5 DALY DW ECHA EDC EEA EFSA EFTA E-PRTR EU EUR FCM GAC GenX GHS HFC KEMI MCL MS NATO NGO NHANES NOx 6:2 Fluorotelomer sulfonate Aqueous film-forming foam (also aqueous firefighting foam) US Agency for Toxic Substances and Disease Register Beauty (or Blemish) Balm / Colour Corrector California Department of Toxic Substances Control Chemical Abstracts Service Harmonised classification and labelling Classification, labelling and packaging or Regulation (EC) No 1272/2008 on the classification, labelling and packaging of substances and mixtures Carcinogenic, mutagenic and toxic for reproduction Alternative name for PFOA (due to its eight carbon atoms) Octamethylcyclotetrasiloxane (D4); decamethylcyclopentasiloxane (D5) Disability-adjusted life year Drinking water European Chemicals Agency Endocrine disrupting chemical/s European Economic Area countries European Food Safety Authority European Free Trade Agreement European Pollutant Release and Transfer Registry European Union Official currency for 19 of the 28 members of the European Union (EU) Food contact material Granular activated carbon Replacement for PFOA Globally Harmonized System of Classification and Labelling of Chemicals Highly Fluorinated Chemical Swedish Chemicals Agency Maximum contaminant level Member State North Atlantic Treaty Organization Non-governmental organization National Health and Nutrition Examination Survey (US) Nitrogen oxide The cost of inaction linked to PFAS exposure 19 NO2 OECD PBT PFAS or PFASs PFBA PFBS PFCAs PFCs PFDA PFDeA PFDoDA PFNA PFHpA PFHpS PFHxA PFHxS PFHxSF PFOA PFOS PFPE PFPeA PFSAs PFTDA PFTrDA PFUnDA PM POPs POSF PPP PTFE PVDF RAC REACH RIVM RME SEAC SEK SMEs SMR SOx SO2 SVHC TDI Nitrogen dioxide Organisation for Economic Co-operation and Development Persistent, bioaccumulative and toxic Per- and polyfluoroalkyl substances Perfluorobutanoic acid Perfluorobutane sulfonic acid Perfluorinated carboxylic acids Perfluorinated compounds Perfluorodecanoic acid Perfluorodecanoic acid Perfluorododecanoic acid Perfluorononanoic acid Perfluoroheptanoic acid Perfluoroheptane sulfonic acid Perfluorohexanoic acid Perfluorohexane sulfonic acid Perfluorohexane sulfonyl fluoride Perfluorooctanoic acid Perfluorooctane sulfonic acid Perfluoropolyether Perfluoropentanoic acid Perfluoroalkane sulfonates Perfluorotetradecanoic acid Perfluorotridecanoic acid Perfluoroundecanoic acid Particulate matter Persistent Organic Pollutants Perfluorooctane sulfonyl fluoride Purchasing power parity Polytetrafluoroethylene (Teflon) Polyvinylidene fluoride Risk Assessment Committee (under REACH) Regulation (EC) No 1907/2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals Dutch National Institute for Public Health and the Environment Risk management evaluation Socio-Economic Assessment Committee (under REACH) Swedish krona Small and Medium Enterprises Standardized mortality ratio Sulphur oxide Sulphur dioxide Substances of very high concern Tolerable daily intake 20 The cost of inaction linked to PFAS exposure TOF UBA UNEP USD USEPA USFAA USFDA VAT VOCs vP vPvB WHO WTP WWTP Total organic fluorine German Federal Environmental Agency (Umweltbundesamt) United Nations Environment Programme United States Dollar US Environmental Protection Agency US Federal Aviation Agency US Food and Drug Administration Value-added tax Volatile organic compounds Very persistent Very persistent, very bio-accumulative World Health Organization Willingness to pay Wastewater treatment plant The cost of inaction linked to PFAS exposure 21 22 The cost of inaction linked to PFAS exposure 1. Introduction Per- and polyfluoroalkyl substances (PFAS) are a large group of chemical compounds that have been used in a wide range of commercial products since the 1950s. They are now found in the environment all around the world. Most people in industrialised countries have one or more PFAS in their blood. PFAS are highly persistent. Though some PFAS may partially degrade under environmental conditions, they will all eventually transform into highly stable end products that will remain in the environment for hundreds or thousands of years2, such that human and environmental exposure will continue long into the future. Human epidemiological studies have found associations between exposure to PFAS and hepatocellular damage affecting liver function in adults, obesogenic effects in females, kidney cancer, low birthweight, reduced length of gestation, and reduced immune response to routine childhood immunizations.3 Because of their persistence, PFAS can travel long distances and have been found even in remote regions such as the high Himalayas and the Arctic where no direct sources of PFAS are known. The compound PFOA, for example, has been found in top predators such as polar bears.4 Moreover, the PFAS tend to be highly mobile and to move readily into ground and surface waters once released to the environment. In the 1950s, when highly fluorinated compounds were first commercialised, the focus was on long-chain PFAS - the so-called C8 substances used in the manufacture of Teflon-coated cookware, water- and stain-resistant textiles, and fire-fighting foams. Evidence emerged in the 1980s and 1990s of the toxicity and bio-accumulability of the longchain PFAS, such as PFOS and PFOA. These long-chain surfactants have been well-studied and are now regulated in different parts of the world to varying extent, leading to complete or partial phase-outs in the EU and the USA. However, PFOA and its derivatives continue to be manufactured in China, India and Russia and as of 2017, China was reported to be the only known manufacturer of PFOS and its derivatives.5 Despite being heavily restricted, these substances are still detected in some consumer products (see section 4.4.3 of this report concerning cosmetics), and other long-chain PFAS continue to be manufactured and used. Some producers have replaced the C8s with shortchain homologues - the C6s and C4s; they claim that the short-chain PFAS are "safer" in 2 Wang Z et al. (2017) A never-ending story of per- and polyfluoroalkyl substances (PFASs)?. Environmental Science & Technology, Mar 7;51 (5). pp 2508-2518. 3 Grandjean P et al. (2014). Changing interpretation of human health risks from perfluorinated compounds. Public health reports, vol. 129: (6). pp. 482-485. 4 Vierke L et al. (2012). Perfluorooctanic acid (PFOA) - main concerns and regulatory developments in Europe from an environmental point of view. Environmental Sciences Europe. v 24: (16). 5 Interstate Technology Regulatory Council (2017). History and use of per- and polyfluoroalkyl substances. The cost of inaction linked to PFAS exposure 23 that they are not as bioaccumulative as the long-chain PFAS. In the meantime, evidence is emerging that the short-chain alternatives pose similar risks to human health.6 Moreover, the number of different PFAS on the global market keeps growing. A 2015 study reported more than 3,000 PFAS were on the global market for commercial use.7 This number was updated in 2018 by a search carried out for the OECD which found over 4,700 different CAS numbers for perfluorinated compounds.8 Other compounds may also be under production, but their identities are protected for confidential business reasons. The number of possible applications of PFAS are also growing rapidly. Figure 1 shows an increasing trend in the number of patents with "perfluor" in the patent text that are approved in the USA each month.9 Figure 1: Number of approved patents in US with "perfluor" in the patent text Source Fischer, S., 2017. "Known uses of PFAS", presentation at Nordic workshop on joint strategies for PFAS, 5.04.2017. 6 Kotthoff M et al. (2015). Perfluoroalkyl and polyfluoroalkyl substances in consumer products. Environmental Science and Pollution Research International. 22(19): 14546-14559. 7 Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. 8 For a list of 4,730 PFAS-related CAS numbers compiled from publicly accessible sources of information, see OECD (2018). Toward a new comprehensive global database of per and polyfluoroalkyl substances (PFASs): summary report on updating the OECD 2007 list of per and polyfluoroalkyl substances (PFASs). 9 Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. Report 7/15. 24 The cost of inaction linked to PFAS exposure A large proportion of these compounds are polymers and therefore exempted from registration requirements under the EU REACH Regulation10; of the others only a few are registered. Very little information is available on quantities produced and for half of all PFAS, almost no information can be found concerning their uses. The quantities of PFAS produced globally also keeps growing. Fluorotelomers used primarily in aqueous firefighting foams (AFFFs), in textiles to provide stain resistance and surface finishing, and as surfactants are a major component of the market. A recent market research report estimated that production of fluorotelomers globally will grow from approximately 21,030,000 kg in 2013 to 47,800,000 kg by 2020, for a 2020 value of USD 539.3 million (EUR 466 million).11 The main drivers of growth are an increased demand from the textile sector (34.8% of total demand in 2013) and government norms leading to use of AFFFs in firefighting systems. Today, PFAS are found in cosmetics, food contact materials, inks, medical devices, mobile phones, pharmaceuticals and textiles. They are used in pesticide formulations, metal production, oil production and mining. They are capable of long-range transport, are highly mobile, and constitute a severe threat to clean water supplies around the globe. The long-term socioeconomic costs of the PFAS already in products or released to the environment are poorly understood. PFAS released over the course of a product's lifecycle will remain in the natural and man-made environments for an indefinite time. One of the concerns is that the contamination may be poorly reversible or even irreversible, and may reach levels that could render natural resources such as soil and water unusable far into the future. This could result in continuous exposure and unavoidable harmful health effects, particularly for vulnerable populations, such as children. For example, PFOS in firefighting foams applied during the 2005 Buncefield explosion contaminated an aquifer that is an important public drinking water source for the Greater London area, so that it is no longer available as a water supply.12 Consensus statements from leading scientists studying PFAS, i.e., the Helsingr Statement13, the Madrid Statement14, and the Zurich Statement15 highlight the health and environmental risks posed by the highly fluorinated chemicals as a group. The statements emphasize the extreme persistence of the carbon-fluorine bond in nature and call for regulatory as well as non-regulatory actions to address the risks associated with all highly fluorinated chemicals, including the short-chain PFAS. 10 Commission Regulation (EU) No 1907/2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC. ("REACH Regulation"), O.J. 396, 30.12.2006, p. 1. 11 Press release, Fluorotelomers Market to Reach USD 539.3 Million Worldwide by 2020, Digital Journal. Accessed 10.11.2018. 12 Matt Gable, UK Environment Agency, as cited in IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 13 Scheringer M et al. (2014). Helsingr Statement on poly- and perfluorinated alkyl substances (PFASs)', Chemosphere, vol. 114. pp. 337-339. 14 Blum A et al. (2015). The Madrid Statement on Poly- and Perfluoroalkyl Substances (PFASs). Environmental health perspectives. Vol. 123, no. 5. pp. A107-11. 15 Ritscher A et al. (2018). Zrich Statement on Future Actions on Per- and Polyfluoroalkyl Substances (PFASs), Environmental Health Perspectives, vol. 126, no 8. The cost of inaction linked to PFAS exposure 25 This study looks at how the production of PFAS, manufacture and use of PFAS-containing products, and end-of-life disposal of PFAS has resulted in widespread environmental contamination and human exposure, resulting in significant socioeconomic costs. It sets forth a methodological framework for estimating costs for society related to negative impacts on the environment and human health, including healthrelated costs and costs for remediation, and uses case studies to illustrate the main impact pathways from PFAS releases and to gather information on direct costs incurred by society to date to reduce exposure to PFAS. The focus of the study is on the costs of inaction in the countries comprising the European Economic Area (EEA). It uses data specific to Nordic countries when available, but also draws cost data from other European countries, the USA and Australia, where relevant. The scope is C4-14 non-polymer fluorosurfactants. It is important to remember that the burden of PFAS-related costs such as healthrelated and remediation costs is largely shouldered by governments and the citizens who pay taxes, while the pollution partly is caused by private operators. By compiling information on societal costs related to PFAS, it is hoped that this study will bring about more effective and cost-efficient management of the risks posed by PFAS. 26 The cost of inaction linked to PFAS exposure 2. The regulatory framework as baseline in relation to PFAS For the purposes of this study, we are defining the "cost of inaction" as the costs to society from existing and future exposures to PFAS if no further measures to curb such exposures are taken. The term "further measures" could refer to additional policy measures as well as better enforcement and implementation of existing policies and regulations.16 The case studies and other information collected for this study are intended to provide an overview of the baseline with respect to PFAS exposure. The aim is dual: (1) to establish a framework for estimating costs for society due to negative impacts on human health and the environment related to PFAS exposure; and (2) to provide monetary values for the costs borne by society, by using costs derived from actual cases involving health impacts or where remedial measures were taken to address PFAS contamination. The overall intention is to highlight the economic case for taking effective and timely action to manage the risks of negative impacts from PFAS exposure. Costs of inaction may refer to different things. One type of cost is related to staying within regulatory guidelines for drinking water (see the subsection below). For example, cases where drinking water supplies were contaminated have led to costs ranging from replacement of water supplies (bottled water, drilling of new wells) to removal of the PFAS contamination from the drinking water by further treatment (reverse osmosis, activated charcoal filters) before delivery to consumers. Another type of cost is the health-related expenses incurred by people exposed to PFAS and suffering from negative health effects as a result. Cases where human populations have been exposed to PFAS over time have been linked to a number of adverse health effects, leading to greater health care costs, loss of production due to absence from work or lower productivity, and a lower quality of life. Less tangible costs might be the loss of use of a natural resource such as groundwater or the loss of property value for homeowners in affected areas. The extreme persistence and mobility in the environment of PFAS is also a consideration, since PFAS contamination tends to continue to spread and costs of clean up through remediation of soil or water will increase if actions are delayed. In recent years, other studies have aimed to estimate costs of inaction related to chemicals exposure. A 2013 UNEP study on costs of inaction on the sound management of chemicals.17looked at available literature concerning environmental and health costs 16 The OECD defines inaction as the lack of development of "no new policies beyond those which currently exist". See OECD. (2008). Environmental Outlook to 2030. Chapter 18: "Chemicals". 17 UNEP (2013). Cost of inaction on the sound management of chemicals. Report Number: DTI/1551/GE. The cost of inaction linked to PFAS exposure 27 linked to a wide range of chemical effects, including heavy metals (mercury, lead), outdoor pollutants (NOx, NO2, PM, SOx, SO2, VOCs), pharmaceuticals and pesticides. Based on data available it is estimated that accumulated health costs related to pesticide poisonings in Sub-Saharan Africa will reach around USD 97 billion by 2020. The 2014 study for the Nordic Council on costs linked to effects of endocrine disrupting substances on male reproductive health is more focused. It reviewed the strength of the evidence regarding negative effects of chemicals considered endocrine disruptors and estimated numbers of incidences of negative effects as well as related costs to society.18 It equated costs of illness with the economic value of reducing risks of exposure to endocrine disruptors. A theme in both studies is the lack of data concerning numbers of chemical exposures and related costs. 2.1 Guideline values for protection of health related to PFAS exposure For the purpose of estimating costs of inaction, it is important to note when levels of contamination require remedial action. Among the tools used by regulatory authorities to control pollutants in environmental media such as groundwater and soil or in water or food for human consumption are limit or guideline values. Such values are important for determining when contamination is at levels that pose unacceptable risks to human health or the environment so that (1) action to remediate the resource is required; and (2) restriction of a certain use or substance is needed to prevent further contamination. Guideline values for acceptable concentrations of PFAS in drinking water are currently in flux. Recent analyses of epidemiological evidence, including of immunotoxicological impairment at background levels of exposure to PFAS19, have led to several regulatory authorities issuing opinions suggesting recommended concentration levels be lower than levels set previously. Most limit values or guidelines to date are for individual long-chain PFAS (PFOS, PFOA, PFHxS, due to their known toxicity and bioaccumulability, e.g., the 2015 World Health Organization recommendation of 0.4 g/l (400 ng/l) for PFOS and 4 g/l (400 ng/l) for PFOA in drinking water. More recent guidelines recognise the potential for harmful impacts from groupings of PFAS, including the short-chain PFAS. This is reflected in the group parameter for PFAS proposed in February 2018 for revision of Council Directive 98/83/EC on the quality of water intended for human consumption (Drinking Water Directive).20 The Commission proposal suggests regulating the whole class of PFAS, i.e., values of 18 Nordic Council of Ministers (2014). The Cost of Inaction - a socioeconomic analysis of costs linked to effects of endocrine disrupting substances on male reproductive health. TemaNord 2014:557. 19 Grandjean P (2018). Delayed discovery, dissemination, and decisions on intervention in environmental health: a case study on immunotoxicity of perfluorinated alkylate substances. Environmental Health (2018) 17:62. 20 Proposal for a Directive of the European Parliament and of the Council on the quality of water intended for human consumption COM/2017/0753 final - 2017/0332 (COD). 28 The cost of inaction linked to PFAS exposure 0.1 g/l (100 ng/l) for individual PFAS and 0.5 g/l (500 ng/l) for PFAS as a group.21 This is an approach already used for pesticides in drinking water. The limit values for PFAS in drinking water set by Sweden and Denmark are also parameters for groups of PFAS. The Swedish National Food Agency has set a group limit value for PFAS at 90 ng/l.22 This also serves as an action level. If the sum of 11 PFAS in drinking water exceeds that level, action is to be taken as soon as possible to reduce the PFAS to concentrations as low as practically possible below that action level. Denmark applies a limit value of 100 ng/l for the sum of 12 PFAS in drinking water (the parameter for PFAS in soil is 0.4 mg/kg TS).23 Germany's Federal Umweltbundesamt (UBA) first published recommended values in 2006 based on a request by the Hochsauerla Valley (see Case Study 3.5.2.2). Since then, new data has led to further revisions and the nd Public Health Department prompted by the PFAS contamination incident in Moehne current UBA guidelines set the lifelong precautionary value at 100 ng/l per se for PFOA and PFOS and 300 when both are present.24 In December 2018, European Food Safety Authority (EFSA) published a scientific opinion on health risks related to PFOS and PFOA in the food chain.25 A previous opinion issued in 2008 set values for tolerable daily intake (TDI) of PFOS at 150 ng/kg bw/day and for PFOA at 1500 ng/kg bw/day. This has been calculated as equivalent to limit values of 70 ng/l for PFOS and 700 ng/l for PFOA. The most recent EFSA opinion sets tolerable daily intake (TDI) for PFOS in food at 13 ng/kg bw/week and for PFOA at 6 ng/kg bw/week.26 This has been calculated as equivalent to limit values of 6.5 ng/l for PFOS and 3 ng/l for PFOA27 which enables the values to be compared to those set for drinking water. In the USA, guideline values are also undergoing revision. In 2016 the US Environmental Protection Agency issued a lifetime drinking water health advisory that set limit values for PFOA at 70 ng/l and for PFOS also at 70 ng/l.28 The advisory notes that when these two chemicals co-occur in a drinking water source, a conservative and health-protective approach would be to set the sum of the concentrations ([PFOA] + [PFOS]) at 70 ng/l. 21 The Commission's explanatory document points out that these values exceed those referred to in Sweden or the USA and therefore compliance should be feasible. 22 Swedish National Food Agency (2017). Riskhantering - PFASs i dricksvatten och fisk. 23 Danish Environmental Protection Agency (2015). Perfluorerede alkylsyreforbindelser (PFAS-forbindelser) incl. PFOA, PFOS og PFOSA. 24 German Environment Agency (2017). Fortschreibung der vorlufigen Bewertung von per- und polyfluorierten Chemikalien (PFC) im Trinkwasser and German Environment Agency (2011). Grenzwerte, Leitwerte, Orientierungswerte, Manahmenwerte Aktuelle Definitionen und Hchstwerte. 25 EFSA Panel on Contaminants in the Food Chain), Knutsen HK et al., 2018. Scientific opinion on the risk to human health related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. EFSA Journal 16(12):5194. DOI: https://doi.org/10.2903/j.efsa.2018.5194 26 From the 2018 EFSA draft abstract on human epidemiological studies. The panel noted that for both compounds exposure of a considerable proportion of the population exceeds the proposed TWIs. 27 Grandjean P (2018). Delayed discovery, dissemination, and decisions on intervention in environmental health: a case study on immunotoxicity of perfluorinated alkylate substances. Environmental Health (2018) 17:62. 28 Environment Protection Agency (2016). Fact Sheet PFOA& PFOS Drinking Water Health Advisories. The cost of inaction linked to PFAS exposure 29 In 2018, the US Agency for Toxic Substances & Disease Register (ATSDR) issued a draft toxicological profile for perfluoroalkyls.29 The draft profile suggested provisional minimal risk levels (MRLs) of 7 ng/l for PFOS and 11 ng/l for PFOA - parameters that are seven to ten times lower than the lifetime advisory levels set by USEPA. Table 5: Regulatory parameters for PFAS in drinking water (DW) (ng/l) Standard PFOS PFOA PFNA WHO guidelines for drinking water (2015) 40 400 Sweden NFA action level (sum of 11 PFAS, 2014) Denmark (sum of 12 PFAS, 2015) Germany (2017) 100 100 60 EU proposed level single PFAS in DW (2018) EU proposed level total PFAS in DW (2018) EFSA TDI in food (2008) 70 700 Draft EFSA TDI in food (2018)4 6.5 3 US EPA lifetime DW health advisory (2016) 70 70 US ATSDR draft finding (2018) 7 11 State of New Jersey (2018) 13 14 13 (binding) State of Minnesota (2017) 27 35 PFAS (single) 100 PFAS (group) 90 1001 3002; 7,0003 500 70 70 Note: 1) Sum of 12 PFAS. 2) PFOS+PFOA. 3) PFAS except PFOS and PFOA. Source: Estimated. Grandjean P (2018). Delayed discovery, dissemination, and decisions on intervention in environmental health: a case study on immunotoxicity of perfluorinated alkylate substances. Environmental Health (2018) 17:62. Several individual US states are setting parameters for PFAS in drinking water at even more stringent levels. In July 2018, the US state of New Jersey adopted a maximum contaminant level (MCL) for perfluorononanoic acid (PFNA) of 0.013 g/l (13 ng/l).30 It is considering the recommendation of the New Jersey Drinking Water Quality Institute to set an MCL for PFOS at 0.014 g/l (14 ng/l). Likewise, the state of Minnesota decided in 2017 to update their health values basing them on the vulnerability of foetuses and infants who are exposed via their mothers, rendering the values significantly lower than those set by the federal USEPA (see Table 5).31 The lowering of mandatory and advisory levels for PFAS in drinking water indicate a growing awareness that exposure to PFAS even at low levels can have negative impacts on human health. In particular, studies have found impaired immunological responses to vaccines at levels of exposure as low as 1 ng/l in serum - levels that are exceeded in most humans.32 29 Agency for Toxic Substances and Disease Registry (2018). Draft toxicological profile for perfluoroalkyls. 30 New Jersey Register, Adopted Amendments: N.J.A.C. 7:9E-2.1; 7:10-5.2, 5.3, and 12.30; and 7:18-6.4. 31 Minnesota Department of Health Perfluoroalkyl Substances (PFAS). Accessed 09.10. 2018. 32 Grandjean P (2018). Delayed discovery, dissemination, and decisions on intervention in environmental health: a case study on immunotoxicity of perfluorinated alkylate substances. Environmental Health (2018) 17:62. 30 The cost of inaction linked to PFAS exposure As the proposal for revision of the EU Drinking Water Directive notes, these substances do not belong in the environment. The proposal points out that Directive 2008/105/EC on environmental quality standards in the field of water policy sets a limit value of 0.65 ng/l for PFOS and suggests a precautionary approach as the way forward. Given that these regulatory parameters are currently a moving target, this study proposes to use Sweden's action level of 90 ng/l as the point of comparison in considering when a resource is considered contaminated by PFAS, such that remedial action should be taken. 2.2 Other regulatory actions underway Other regulatory efforts underway are aimed at controlling PFAS on the market, because of evidence of their negative impacts. Within the European Economic Area (EEA), member countries are subject to the provisions of the EU REACH Regulation, as well as to the regulation implementing the Stockholm Convention on persistent organic pollutants. PFOS has been restricted under the Stockholm Convention since 2009. During the fall of 2014, Norway and Germany joined in submitting a proposal for the EU to restrict PFOA, its salts and related substances.33 This led to the adoption of Commission Regulation (EU) 2017/1000 of 13 June 2017 amending Annex XVII to REACH, as regards perfluorooctanoic acid (PFOA), its salts and PFOA-related substances. In March 2017, Sweden and Germany proposed to consider PFHxS a substance of very high concern.34 This was adopted by the European Chemicals Agency (ECHA) later the same year, and the substance is now on the Candidate List. Norway has registered an intention to submit a restriction proposal for PFHxS under REACH. Sweden and Germany also jointly proposed in 2017 to restrict the manufacturing and placing on the market of six PFAS (PFNA, PFDA, PFUnDA, PFDoDA, PFTrDA and PFTeDA), as well as their salts and precursors.35 The aim in restricting these long-chain (C9-C14) PFAS is to prevent industry from switching to them once the restriction of PFOA goes into effect in 2020. Both the RAC (Risk Assessment Committee) and the SEAC (Committee for Socio-economic Analysis) have agreed to the restriction proposal; public consultation on the SEAC opinion closed on 19 November 2018. The Stockholm Convention on persistent organic pollutants (POPs) continues to consider measures related to PFAS additional to the 2009 listing of PFOA for global restriction (Annex B). In September 2018, the POPs Review Committee agreed to recommend to the Parties to the Convention that PFOA be phased out, because its PBT qualities, the occurrence of PFOA in environmental compartments, and the evidence of long-range environmental transport supported the conclusion that it is likely to lead to significant adverse effects such that global action is warranted. It also evaluated the 33 ECHA (2014). Germany and Norway propose a restriction on Perfluorooctanoic acid (PFOA), its salts and PFOA related substances. 34 ECHA (2017a). Inclusion of substances of very high concern in the Candidate List for eventual inclusion in Annex XIV. 35 ECHA (2017b). Public consultation. Germany, in collaboration with Sweden, proposed a restriction on C9-C14 perfluorocarboxylic acids (PFCAS), their salts and related substances (precursors). The cost of inaction linked to PFAS exposure 31 exempted uses of PFOS based on the availability of alternatives and recommended most of them for removal or to be made time-restricted. The POPs Review Committee also adopted the risk profile for PFHxS, thereby moving it to the next stage of a risk management evaluation (part of the process for considering whether to list a chemical in the Convention).36 The next meeting of the Parties takes place in April 2019 when the decisions will be taken on the table concerning the listing of PFOA for global phase-out and for removing exemptions for uses of PFOS. 36 Stockholm Convention (Website) Report of the POPs Review Committee at the work of its fourteenth meeting, UNEP/POPS/POPRC.14/6. 32 The cost of inaction linked to PFAS exposure 3. Methodology to assess environment and health-related costs This chapter describes the methodology for building an integrated socioeconomic model to assess the environmental and health-related costs of PFAS exposure in European countries. No such methodology had been developed at the time this study was conducted. In a stocktake of socioeconomic assessments for PFOA and its salts carried out for the OECD37, the need for a method to draw together information on the longterm environmental and health costs related to PFAS exposure was expressed. The impact pathway shown in Figure 2 provides an overall framework for the socioeconomic analysis. Figure 2: Generic impact pathways for linking substances to possible impacts 37 Gabbert, S. (2018). Economic assessments and valuations of environmental and health impacts caused by Perfluorooctanoic acid (PFOA) and its salts. OECD Environment Working Paper No 128. The cost of inaction linked to PFAS exposure 33 The figure links production and use of PFAS to impacts and their economic valuation. It provides a template for assessment of each source, enabling the analyst to consider which impacts are relevant in a case. Following on from Figure 2, data needs for the assessment are identified in Table 6, which presents an overview of the socioeconomic assessment method that guided the analysis. The method defines six stages for the assessment of both health and environmental impacts. For each stage, it defines a set of input parameters, data sources and key assumptions by stage of the assessment process. The evidence available on the extent of PFAS exposure and its impact is often restricted to a case of contamination in a specific geographic area. One of the key assumptions noted for several stages is therefore the transferability of scientific findings from one specific context to another. Table 6: Assessing the socio-economic impacts of PFAS exposure Stage Input parameters Data sources Key assumptions Defining sources of PFAS exposure Identification of impacts Quantification of impacts Valuation Value transfer Aggregation Uses of PFAS, e.g., production, product manufacture, product use Number of activities involving PFAS (by use) Identification of impact pathways (by use) Health impacts Listing of impacts linked to PFAS, e.g. cancers Environmental impacts Contamination of resources such as drinking water For each effect identified above: Size of population or receiving body at risk Prevalence of disease Response function Health Unit values e.g. EUR/death Environment Willingness to pay to avoid loss of ecosystem services Cost of damage to commercial fisheries, agriculture, etc. Cost of environmental remediation Factors including exchange rates, size of population affected and income levels to improve applicability of values to the target population Discount rates Contextual data permitting quantification of effects beyond the case study materials that are available Scientific and grey literature, on-line research Case studies Case studies Case studies and scientific literature National (etc.) statistics, scientific literature, including reviews Documents submitted to ECHA, OECD, etc. Further review of the literature Documents submitted to ECHA. OECD, etc. (eg. D4/D5 dossier) Further review of the literature. Market prices Published case study materials Valuation literature, exchange rate databases Standard European Commission practice (constant 4%) + alternatives of 0% and 2% National (etc.) statistics, scientific literature, including reviews Future applications Future use, alternatives Assumptions of future conformance Causality for PFAS in general and then for specific PFAS Transferability Transferability Transferability Range of factors that should be accounted for Validity of constant rates over extended timescales Transferability 34 The cost of inaction linked to PFAS exposure The specificities of the methodologies for assessing health impacts can be found in Section 5.1 while those for the environmental impacts can be found in Section 5.2. The lack of systematic and standardised evidence in addition to the underlying uncertainties with regards to the extent of impact and their consequences presented challenges for the process of developing quantitative estimates. The robustness of each quantitative finding is explored through sensitivity analysis. The sensitivity analysis allowed for the construction of lower and upper bounds for each quantitative estimate to reflect the underlying uncertainties. Other costs that could not be quantified, but for which strong evidence was identified, are assessed in a qualitative manner. 3.1 Health-related costs This section describes the specific methodology for assessing the health-related costs of PFAS exposure. A growing body of scientific literature suggests that PFAS exposure can lead to a wide range of adverse health impacts at different levels of exposure. To date, a monetisation of these health-related costs has not been developed due to the lack of a global consensus on the specific health impacts linked to PFAS exposure and a complete understanding of the level of exposure needed to trigger a health impact. This cost of inaction assessment therefore presents a first attempt to monetise the impacts for several of the identified health endpoints of PFAS exposure. The methodology takes findings from various epidemiological studies showing relative risks due to exposure. It considers "what-if" scenarios, where scenarios assume the transferability of epidemiological studies from one context to another in some cases, and the subsequent impacts. It then extrapolates those findings to the "exposed" population in the Nordic countries and the EU. The design of these scenarios attempts to reflect the underlying uncertainties with respect to the level of exposure and the epidemiological evidence. The methodology for assessing the health-related costs follows four basic steps: (1) identification of endpoints, (2) responses to levels of exposure, (3) quantifying impacts and (4) aggregation. Each step is described below. 3.1.1 Identification of endpoints In assessing the potential health impacts of exposure to PFAS, toxicologists and epidemiologists need information on both human health endpoints (which are defined as conditions or diseases that reflect poorer health and an increased risk of mortality) and substance exposure. Data on human health endpoints may be obtained from for example from public health records or from surveys of exposed individuals. The level of human exposure to PFAS can be inferred from e.g. data collected from monitoring PFAS contamination in drinking water or other local sources, or it can be investigated more closely through the analysis of blood samples. Identifying health impacts related to exposure to PFAS is challenging for several reasons. Health impacts are typically identified through studies that compare a relatively "exposed" group and a relatively "unexposed group" while controlling for other factors The cost of inaction linked to PFAS exposure 35 related to the health endpoint. From a causal perspective the ideal setting would be a randomized control study (experiment), in which individuals would be assigned at random into treatment and control groups (exposed/unexposed groups). This would ensure random assignment of background characteristics in each group, i.e the background characteristics in each group would on average be the same. However, in settings with environmental pollution or toxic substances, this is neither ethical nor economically feasible. A potenitial solution to this metodological problem is to look for a "natural experiment" where the exposed group has similar background characteristics as the unexposed group, for example two neighborhoods in close geographic proximity where one is more exposed than the other. However, given the persistence of PFAS and their ability to travel long distances, it is unlikely that a group in close proximity to the contaminated area would not have an elevated level of exposure. In fact, considering that PFAS have been found in some of the most remote places in the world, it may not be possible to find a truly unexposed group among humans. To further complicate inference, contamination often consists of more than one PFAS compound, making it difficult to attribute the exposure to a single compound. The lack of regular biomonitoring of PFAS in humans (through collection and analysis of blood samples) in many countries also presents a severe limitation in the data available on the health impacts of PFAS at different levels of exposure and in different contexts. In addition, the sample sizes for most epidemiological investigations are quite small, which limits the extent to which health endpoints can be identified with a reasonable level of confidence. This is especially challenging for health endpoints that are relatively uncommon in the general population such as testicular and kidney cancer. In conclusion, the above mentioned reasons may explain why some epidemiological studies find statistically significant effects while others do not. The first part of Annex 2 presents an overview of epidemiological studies that have linked PFAS exposure with a range of different health endpoints. The sample size, the population studied, and the time period are indicated for each study. Several regulatory bodies and expert panels around the globe have carried out reviews of the scientific evidence in order to reach conclusions about the relevant health endpoints of human exposure to PFAS. These reviews usually seek a certain level of consensus across different scientific studies in order to conclude that PFAS have an adverse impact on that health endpoint. In drawing conclusions about specific health endpoints, these reviews utilise a "strength of the evidence" approach. The scope of the evidence considered as well as the evaluation of the strength of the evidence varies, however, across the available reviews. This leads to different conclusions in terms of the recognised health endpoints of PFAS (see Table 7). For example, an Expert Health Panel convened by the Australian Department of Health presented a review of recent literature reviews of the potential health impacts of PFAS in May 2018.38 The study did not find conclusive evidence for any of the health 38 Australia Government Department of Health (2018). Expert Health Panel for Per- and Poly-Fluoroalkyl Substances (PFAS). 36 The cost of inaction linked to PFAS exposure endpoints identified in other expert reviews. However, the Panel came to the conclusion that because current evidence is primarily based on weak study designs and is inconsistent in many respects, some degree of important health effects for individuals exposed to PFAS could not be ruled out based on the existing evidence. The provisional report from the European Food Safety Authority (EFSA) concludes that PFAS exposure can lead to metabolic disease, immunotoxicity, and developmental toxicity, but finds the evidence with respect to cancer and endocrine disruption not sufficiently robust. The draft toxicological profile from the US Agency for Toxic Substances and Disease Registry (ATSDR) on the other hand suggests that the evidence demonstrates a relationship with several health endpoints such as asthma, pregnancy-induced hypertension and an increased risk of thyroid disease. This assessment applies not only for PFOA and PFOS, but also several other PFAS compounds such as PFHxS and PFDeA. Table 7: Reviews of health endpoints linked to PFAS exposure Category Health endpoint EFSA Metabolic disease Liver damage PFOA) Ulcerative colitis Increased serum cholesterol levels (PFOS, PFOA) Immuno- toxicity Endocrine disruption Decreased immune response (e.g. antibody response to vaccines) Increased risk of asthma diagnosis Increased risk of thyroid disease (elevated hormones) Elevated sex hormones Decreased fertility (PFOS) Developmental outcomes Carcinogenicity Pregnancy-induced hypertension/pre-eclampsia Delayed menstruation and earlier menopause Lower birth weight Testicular and kidney cancer (PFOS, PFOA) ATSDR (PFOA, PFOS, PFHxS) (PFOA, PFOS, PFNA, PFDeA) (PFOA, PFOS, PFHxS, PFDeA) (PFOA) C8 Health Project (PFOA, PFOS) (PFOA, PFOS) (PFOA, PFOS) (PFOA, PFOS) US EPA (PFOA) (PFOA, PFOS) (PFOA) (PFOA) (PFOS) (PFOA) (PFOA) (PFOS) (PFOA) OECD Another key assessment was made by the C8 Health Project, which was established as part of a class action legal settlement made by a chemical manufacturer in West Virginia.39 The C8 Health Project was led by a Science Panel of three epidemiologists who carried out a series of scientific studies using biomonitoring data gathered from the site. 39 The chemical manufacturer was Dupont. In 2017, Dupont merged with the Dow Chemical Company. Dupont is now a subsidiary of the Dow Chemical Company. The cost of inaction linked to PFAS exposure 37 The Science Panel concluded that the exposure had a probable link with seven health conditions and diseases: High cholesterol (hypercholesteremia), ulcerative colitis, thyroid function, testicular cancer, kidney cancer, preeclampsia, as well as elevated blood pressure during pregnancy (pregnancy-induced hypertension). These endpoints roughly reflect the health impacts identified in a 2018 review conducted for the OECD of epidemiological research studies.40 Lastly, the US EPA issued two reports in 2016 for the health effects for PFOA and PFOS.41 It recognised almost all of the health endpoints identified by other reviews and panels. Some epidemiological studies were considered as conclusive evidence in support of a certain health endpoint by one assessment, but not as conclusive by another. For example, the EFSA opinion prominently cites a study conducted as part of the C8 Health Project42 as providing strong evidence that low birthweight is a relevant health endpoint.43 Yet the C8 Science Panel did not find the evidence sufficiently strong for low birthweight as an endpoint. The lack of consistency across these assessments and the differential weighting of specific epidemiological studies reflects the challenges at present to reach a global consensus on the health endpoints of PFAS contamination. As per Table 7, the body of scientific evidence and research has grown over time, collectively suggesting that PFAS do have adverse health impacts on humans. A July 2018 study44 notes that "[a]ccumulated evidence from studies of experimental animal models and of humans from highly exposed populations supports the conclusion that PFOA and PFOS, along with other carboxylate and sulfonate PFAS, are multi-system toxicants. In other words, exposure to PFAS is associated with toxicological findings in many types of tissues and systems." The International Agency for Research on Cancer (IARC) provides additional insights about the carcinogenicity endpoint. The agency published a monograph in June 2018 concluding that there is limited evidence in humans for the carcinogenicity of PFOA. However, a positive association between PFAS exposure and cancers of the testis and the kidney was observed, rendering the overall evaluation that PFOA is possibly carcinogenic to humans.45 This cost of inaction assessment of PFAS is built on the assumption that exposure to PFAS is linked to the health endpoints indicated in Table 7. In addition, the assessment assumes that different short- and long-chain PFAS compounds will lead to similar adverse health impacts. 40 Gabbert S (2018). Economic assessments and valuations of environmental and health impacts caused by Perfluorooctanoic acid (PFOA) and its salts. OECD Environment Working Paper No 128. 41 EPA (2016). Health effects support document for perfluorooctanoic acid (PFOA). Document no: EPA822R16003 and Health effects support document for perfluorooctane sulphate (PFOS). Document No: EPA822R16002. 42 Stein C R et al. (2009). Serum levels of perfluorooctanoic acid and perfluorooctane sulfonate and pregnancy outcome. American journal of epidemiology, 170(7), pp.837-846. 43 The main reason for its prominence is its large sample size. The authors did not consider other studies investigating this endpoint to be sufficiently powered. 44 Hopkins et al. (2018). Recently Detected Drinking Water Contaminants: GenX and Other Per- and Polyfluoroalkyl Ether Acids. American Water Works Association Journal 110:7 p.13-28. 45 IARC (2018). Perfluorooctanoic acid (PFOA). 38 The cost of inaction linked to PFAS exposure 3.1.2 Responses to levels of exposure In the absence of data showing clear dose-response relationships, this study has instead considered health impacts found by epidemiological studies of populations at three levels of exposure: (1) background (low) PFAS exposure levels; (2) elevated (medium) PFAS exposure levels and (3) occupational (high) exposure. Individuals with a higher level of exposure to PFAS can be expected to have a higher concentration of the contaminant in their blood, resulting in elevated health risks. The Expert Health Panel from Australia noted that individuals in highly exposed communities typically have a blood serum concentration about ten times higher than the general population while workers can have a blood serum concentration one thousand times higher than the general population.46 In other words, it is assumed that the blood serum concentration serves as a rough proxy of PFAS exposure. Blood samples from population studies such as the National Health and Nutrition Examination Survey (NHANES) in the United States were assumed to provide information on background (low) PFAS exposure. Medium and high PFAS exposure levels were drawn from epidemiological studies based on populations affected by known incidents of PFAS contamination. Communities in close proximity to PFAS production or with PFAS-contaminated drinking water were considered to have been exposed at medium levels while occupationally exposed individuals, e.g., at chemical production plants, were considered to have high levels of exposure. Some adverse health impacts appear to materialise at high PFAS exposure levels while for other conditions background exposure generates an elevated risk of disease. Carcinogenicity is most clearly linked with a high level of exposure to PFAS. One study47 found the risk of kidney cancer among residents in close proximity to PFAS contamination was elevated only among those with an above-average level of exposure. Factors that may lead some residents in the same community to have a higher level of exposure than others include the person's age, the number of years lived in the community, and the level of contamination in his or her household's drinking water. The IARC (2018) assessment also concludes that the evidence regarding the elevated risk of kidney cancer is most convincing for the case of occupational exposure, the highest category of exposure.48 At the same time, it is not possible to rule out that other residents in the same community with a below-average exposure to PFAS do not have an elevated risk of kidney cancer. The lack of statistical significance may be due to the low risk of kidney cancer in the general population.49 46 Australia Government Department of Health (2018). Expert Health Panel for Per- and Poly-Fluoroalkyl Substances (PFAS). 47 Vieira V M et al. (2013). Perfluorooctanoic acid exposure and cancer outcomes in a contaminated community: a geographic analysis. Environmental health perspectives, 121(3). 48 IARC (2018). Perfluorooctanoic acid (PFOA). 49 National Cancer Institute [NCI] Surveillance, Epidemiology, and End Results data for 2005-2011. The kidney cancer endpoint is also difficult to assess as the survival rate is high at 73%. Testicular cancer, which is another possible endpoint of PFAS exposure, has an even higher survival rate of 95%. Studies with a large sample over a long time frame have a greater likelihood of detecting a differential risk of developing these types of cancer. The cost of inaction linked to PFAS exposure 39 Metabolic disease is linked with all levels of PFAS exposure with less severe endpoints such as elevated cholesterol and ulceritis associated with background (low) exposure and more critical endpoints such as liver damage associated with elevated (medium) exposure. The link between the level of PFAS exposure and health endpoints appears to be similar for endocrine disruption and developmental toxicity. Studies suggest that background (low) levels of PFAS contamination can lead to immunotoxicity, especially for infants and children. 3.1.3 Quantifying impacts A quantitative valuation or monetisation was then undertaken for a selection of the identified health endpoints for which there is a reasonable level of global consensus. Given the limited epidemiological evidence from the specific contexts, the assessment considers "what-if" scenarios. These scenarios assume the transferability of quantitative findings from epidemiological studies from one context to another in some cases, and the subsequent impacts. The quantitative findings from epidemiological studies are typically in the form of a standardised mortality ratio, an odds-ratio or a relative risk. These statistical terms reflect the elevated risk in an "exposed" population compared with a "less exposed" population after controlling for other factors.50 The assessment selected risks that were estimated with a high level of confidence. In statistical terms, this implies that the risk was estimated with a margin of error of 5% or less. The what-if scenarios constructed define linkages between PFAS exposure, the health endpoint and mortality related to the endpoint. Additional studies that investigated the relationship between the health endpoint in question and mortality were also reviewed to characterise the last linkage in the chain of the what-if scenario. The estimated number of health impacts such as deaths through the what-if scenario was then monetised using the "value-of-statistical-life" approach. The quantitative analysis applies the lower bound of the range recommended by ECHA of EUR 3.5 to EUR 5 million per life lost.51 The ECHA value of life estimates are commonly used by regulatory agencies in Europe. The lower bound value is also comparable to the value of a statistical life in the EU reported by the OECD, which was EUR 3 million in 2012.52 When adjusted for inflation, the OECD value approaches the minimum ECHA value. The monetised figures presented in this study should be understood as the minimum health-related costs of inaction in that only a few health endpoints and a few impacts of these endpoints could be investigated. More data on exposure-response relationships is 50 Epidemiological models may control for a wide range of individual and environmental factors such as age, gender, educational background and occupation. 51 ECHA (2016). Valuing selected health impacts of chemicals: Summary of the results and a critical review of the ECHA study. 52 OECD (2012). Mortality risk valuation in environment, health and transport policies. 40 The cost of inaction linked to PFAS exposure needed for estimating the impacts with respect to other endpoints. The monetised figures can provide an indication of the potential costs of inaction, and where more research is needed to develop more robust what-if scenarios and estimations. 3.1.4 Aggregation Estimates of impacts or costs were then aggregated across the Nordic countries and the EEA when possible. A major challenge in carrying out the aggregations was the determination of the size of the relevant exposed population. For example, the scenarios for occupational (high) and elevated (medium) levels of exposure required assumptions regarding the number of chemical plants or other sources of PFAS contamination and the average number of persons exposed in each location. To the extent possible, information to use as a basis for these assumptions were gathered through the case studies. A range of plausible values was considered in the absence of conclusive information. The scenarios for background (low) exposure assume that all individuals living in the Nordic countries or the EEA are affected. The estimations for background (low) exposure could then draw on country-level population statistics available from Eurostat. The annual number of deaths or cases of a health conditions were generated for the exposed population based on the level of exposure and the epidemiological evidence. The additional level of mortality and disease can therefore be understood as the annual costs due to PFAS exposure for the relevant geographic area - the Nordic countries or the EEA - under the specific scenario. These costs would not be incurred if the PFAS exposure, as defined by the scenario, did not occur. Therefore these estimated costs can be understood as potential costs of inaction. The analysis generated a point estimate, a lower bound and an upper bound for each scenario. The generation of these three different values stem from the findings of the epidemiological studies, which provided the parameter relating to the level of exposure to PFAS and the elevated health risk. The point estimate was calculated using the main finding from the epidemiological study. The 95% confidence interval for the finding from the epidemiological study was used to generate the lower and upper bound estimates. The findings from the estimations can be found in Section 5.1. The calculations underlying these estimations including the parameter values can be found in Annex 2. 3.2 Environment-related costs Two distinct types of environment-related cost are associated with the present research: environmental remediation; and loss of ecosystem services. The cost of inaction linked to PFAS exposure 41 3.2.1 Approaches to environmental remediation The most technically and economically efficient techniques for reducing contamination of the environment with PFAS arise at manufacture and application prior to sale to endusers, and are as follows: using alternatives; improved containment at industrial sites manufacturing PFAS; increased use efficiency of PFAS materials, for example by re-cycling unused solution at manufacturing sites. Recycling of most applications post-consumer is generally not feasible. A possible exception may be AFFFs that have gone past their use-by date; and improved containment at industrial sites using PFAS, for example using controlled application processes and controlled disposal of residues. Using these techniques, PFAS will either not be present at all, in cases where non-PFAS alternatives are used53, or where they are used, the volume of material requiring treatment will be limited. However, these measures address only part of the problem: except for the use of alternatives, they are all associated with some level of discharge and they cannot influence contamination associated with goods used outside manufacturing or processing sites. In some cases, AFFFs being a notable example, dispersion to the environment is immediate and total, upon use. Costing the first set of options set out above is beyond the scope of this report. Costs of improving containment and use efficiency54 will be internalised by producers and business consumers. In many cases any improvement and associated costs will reflect legislative demands and the permit conditions that facilities are required to operate to. There will be some cost to government, for example through the use of PFAS alternatives at military airbases. Analysis of these costs would require a counterfactual scenario to be developed where alternatives to PFAS were defined, along with differences in likely permit conditions, extending the current assessment to a full cost-benefit analysis. Environmental remediation may deal with: contaminated soils contaminated groundwater contaminated surface water targeted collection of goods containing PFAS at end-of-life to reduce the volume of material that needs to be treated. 53 Some "alternatives" are not impact-free. The BREF document on the tanning of hides and skins notes a trend to using shorter chain (C4 or C6, rather than C8 perfluoro compounds). The shorter chain fluorocarbon resins are described (as of 2013) as being more favourably assessed toxicologically but are as persistent in the environment as the longer chain PFAS. 54 Increased use efficiency will lead to at least some payback through reduced demand for new material. 42 The cost of inaction linked to PFAS exposure Targeted collection of goods containing PFAS at end-of-life is not a viable option in most cases. Some applications, such as the use of aqueous film-forming foams (AFFFs) lead to direct contamination of the environment. The range of other applications has become so extensive that separation of goods, for example at waste processing sites, would be impractical, as is evident from the problems faced across Europe with respect to the efficient recycling of another widely dispersed material, plastic. Even if it were practicable to some extent, the option would still leave some level of contamination present as collection processes would never capture all contaminated material and emissions will arise during the use phase. Environmental remediation of PFAS contamination is not straightforward. Concawe55, an environmental research and advisory body for the oil industry, notes that many remediation techniques used for other contaminants are ineffective for PFAS because of their low volatility (preventing the use of gas stripping) and their resistance to biological degradation. With this in mind, the main techniques for removing PFAS from the environment involve: removal of soils; and groundwater extraction and PFAS adsorption onto activated carbon or resins. Soil removal is a reasonably straightforward process, though it can clearly create a large quantity of soil that needs either further treatment or storage. Storage of PFAS at authorised landfills may lead to leaching into surrounding areas because standard leachate treatment plants are not able to effectively treat these substances.56 The use of landfill would also place a burden on space at landfills, particularly if contaminated material is required to be stored at the limited number of sites that are licensed to deal with hazardous wastes. One alternative to storage at landfill is to destroy PFAS through incineration, though this itself is not straightforward (or inexpensive) as it requires use of very high temperatures. Complete destruction of PFOS requires a temperature of 1,000 to 1,200 C.57 Associated costs are high because of the significant energy inputs that are required and the likelihood that soil (etc.) volumes will be large. Modern municipal and hazardous waste incinerators can reach high temperatures, e.g. 800 C or more, and at such temperatures the PFAS on a treated consumer product may break down. The use of lower temperatures can lead to the generation of hazardous by-products.58 The Concawe report also discusses stabilisation of PFAS within soils using additives such as activated carbon, and solidification of soils using concrete mixes. Both have been shown to greatly reduce the potential for leaching. However, neither approach provides 55 Concawe (2016). Environmental fate and effects of poly- and perfluoroalkyl substances (PFAS). Report no: 8/16. 56 Oliaei F et al. (2013). PFOS and PFC releases and associated pollution from a production plant in Minnesota (USA). Environmental Science Pollution Research. 20: 1977-1992. 57 Schultz M et al. (2003). Fluorinated alkyl surfactants. Environmental Engineering Science 20(5) 487-501. Yamada T et al. (2005). Thermal degradation of fluorotelomer treated articles and related materials. Chemosphere 61, 974-984. 58 Yamada T and PH Taylor (2003). Final Report: Laboratory scale thermal degradation of perfluorooctanyl sulfonate and related precursors. Final Report for 3 M Company. The cost of inaction linked to PFAS exposure 43 a permanent solution to the problem of PFAS contamination. These approaches could be seen as creating a potentially large number of hazwaste landfills across Europe. A further option is soil washing, moving PFAS to the aqueous phase where it can be filtered and retrieved. Concawe reports trials showing a reduction in concentrations below target levels after two washing cycles.59 No information was available on the quantities of sludge or filtrate generated, materials that would need further storage or treatment. Concentrating the PFAS up would have the benefit of reducing the quantity of contaminated material, which would in turn reduce demand for storage or incineration. The most commonly applied treatment for contaminated groundwater is extraction and use of granular activated carbon (GAC). The efficiency of extraction of PFOS is in the order of 90%, though efficiency for other PFAS (e.g. PFOA) can be much lower especially for short-chained PFAS like PFBA. The characteristics of the absorptive medium can be adapted to specific PFAS, though this leads to trade-offs with improved recovery of some species and lower recovery of others. Spent recovery media are typically incinerated at high temperature. Other effective treatments are reverse osmosis which is commonly used for preparation of drinking water60, ion exchange and nano-filtration. All cases will generate wastes that require either specialised storage or high temperature incineration. Concawe reports on a number of other innovative methods currently being explored, such as photolysis/ photocatalysis, reductive decomposition, advanced oxidation and sonolysis. However, none seem close to application. In particular, these technologies are unlikely to be feasible for high flowrate, low concentration applications, precisely the conditions faced for environmental remediation. The potential for contamination of surface waters may also need to be considered. In many cases, contamination would be better treated at source (i.e. at the factory or at a site of soil contamination) than downstream where concentrations will be more dilute. Information from the Veneto Region below provides a case where contamination may be from diffuse sources across the Region. A first response should be to ensure that the emitting industries either treat their own waste water or discharge to sewer for treatment at a waste water treatment plant (WWTP) if this does not already happen. This may further require upgrading of the WWTP and where contamination of surface waters persists, further remedial action may be necessary. In summary, a number of techniques are available for PFAS remediation, but they are not straightforward and as such are likely to be costly. The information above has been used to check the validity of options adopted at various locations in the case study material presented below. 59 The report does not state what these target values were. Other parts of the report discuss a range of 0.1 to 0.5 g/l, though elsewhere reference is made to 0.023 g/l. 60 Tang C Y et al. (2007). Effect of Flux (Transmembrane Pressure) and Membrane Properties on Fouling and Rejection of Reverse Osmosis and Nanofiltration Membranes Treating Perfluorooctane Sulfonate Containing Wastewater. Environmental Science and Technology 41: 2008-2014. 44 The cost of inaction linked to PFAS exposure 3.2.2 Quantifying the costs of environmental remediation Building on the information presented in the previous section, the costs to regulatory and other stakeholders (excluding direct impacts on health and the environment) relate to the following activities: survey work to identify sites that are likely not to meet regulatory criteria on PFAS contamination; liaison with stakeholders using or living near contaminated resources (public meetings, etc.); monitoring concentrations of PFAS before, during and after remediation; removal of contaminated soils followed by: storage of these soils at controlled landfill sites, or high temperature incineration, or soil washing to extract PFAS storage or high temperature incineration of contaminated filters, etc. extraction of groundwater followed by: use of GAC to absorb PFAS, or use of ion-exchange (IX) resins to absorb PFAS, or reverse osmosis or nanofiltration, which may require expansion of water treatment facilities given the need to generate an additional 15-20% of water for use in the treatment process61 all of which would need to be followed by storage of collected pollutants or high temperature incineration of contaminated filters, etc. Stabilisation of PFAS within soils is not considered here to be a long-term solution to the problem for the vast majority of cases, and so is not considered further. Cost data for the above activities have been taken from the case study examples provided in Chapter 4 and various sources in the grey literature (see Annex 3 for more on this). Using data from reported cases, analysis can be carried out in two stages: 1. Quantification of the costs of environmental remediation for a standardised situation, e.g. a. release of a standardised quantity of PFAS into soil (e.g. 10kg); and b. contamination of a drinking water supply i. leading to the provision of water from alternative sources for a standardised population (e.g. 100,000 people) 61 Black and Veatch (2018) Alternatives Evaluation Report. Emerging contaminants treatments strategies study. The cost of inaction linked to PFAS exposure 45 ii. leading to the requirement for upgrading of water treatment works for a standardised population (e.g. 100,000 people) 2. Extrapolation of the "standardised cases" to national or European level, accounting for: a. the number of sites where soils are contaminated by PFAS and where soils will need to be decontaminated or stabilised; b. the number of communities served by contaminated water supplies in need of remediation; c. the probability that these sites will be remediated; and d. the population served by contaminated drinking water supplies. The analysis that follows identifies a number of uncertainties that affect the analysis, including: the quantities of PFAS released; the number of contaminated sites; the representativeness of case study material; and Limited data availability for some activities. The extrapolation of existing data to new cases is clearly prone to uncertainty. The easiest way of dealing with this would be to define plausible ranges for each variable used in the extrapolation, and to use these to calculate absolute minimum and maximum estimates of cost. Such an approach has three problems: 1. It can provide an extremely broad spread of results. 2. It biases attention to the extremes of the range rather than to the most likely esti- mate, providing no reason to prefer any value between the two. 3. It typically does not account for possible correlation between input values. Consideration was given to the use of Monte Carlo analysis to deal with uncertainty. However, it was concluded in the course of the work that this could give a false impression of the quality of the data adopted for analysis. An alternative developed for the presentation of results was to adopt a series of plausible and well-defined scenarios to demonstrate the likely and potential magnitude of damage and reasonable ranges for the results. The assumptions followed for each scenario would be clearly stated with the results in order that readers can understand why and how differences arise. 46 The cost of inaction linked to PFAS exposure 3.2.3 Quantifying the costs related to loss of ecosystem services Ecosystem services are typically regarded as falling into four categories: supporting services, such as nutrient cycling and soil formation provisioning services, such as the production of food, fishing opportunity, raw materials, novel compounds of possible medicinal use and clean water regulating services, such as carbon sequestration, water purification, waste cycling, pollination and pest control; and cultural services, such as the spiritual and historic significance of natural resources. Whilst the possibility of PFAS affecting a number of these services is not ruled out, there is a lack of data available for describing associated impact pathways. The most significant omissions concern information on response functions for specific ecosystems and data on stock at risk for specific sites affected by PFAS contamination. An alternative route, proceeding from emission straight to valuation, is possible using results of studies undertaken to quantify individual willingness to pay (WTP) to avoid the impacts of either PFAS or other substances with PBT or vPvB properties. Two examples have been identified: 1. A study by Sunding (2017) assessing damage to Minnesota's natural resources resulting from 3M's disposal of PFCs in Washington County, Minnesota62. 2. A UK survey undertaken not on PFAS specifically but for the REACH restriction dossier for D4/D563 that considered WTP regarding PBT and vPvB properties. Sunding's study from the USA covered damage to groundwater, surface water and increased costs of water purification. The groundwater assessment compared the difference in house prices in areas where groundwater was contaminated and used for drinking water, with areas that were uncontaminated. Results showed a decline in house prices of 7.3% in the most affected areas (4.4% in other contaminated communities). The average lost value per home was USD 17,400 or EUR 12,657 (USD 14,000 or EUR 10,184 in other affected communities) with an annualized loss of USD 288 or EUR 209 (USD 231 or EUR 209). Total past and future damage for the period 1971 to 2050 for the 57,000 houses affected was estimated to USD 1.5 billion, or EUR 1.1 billion. Results were corroborated with surveys of residents and consideration of defensive expenditures via bottled water sales. As part of the assessment WTP to avoid fishing in PFOS contaminated surface waters was estimated from a survey at USD 19 (EUR 14) to USD 45 (EUR 33) per trip depending on the type of fish present ("popular" 62 Sunding D L, Damage to Minnesota's natural resources resulting from 3M's disposal of PFCs in Washington County. 63 RAC/SEAC (2016) Background Document to the Opinion on the Annex XV dossier proposing restrictions on Octamethylcyclotetrasiloxane (D4) and Decamethylcyclopentasiloxane (D5). Committee for Risk Assessment and Committee for Socio-Economic Analysis, European Chemicals Agency, Helsinki. The cost of inaction linked to PFAS exposure 47 or "unpopular" species). Total loss to anglers over the period 2008 to 2040 was estimated at USD 121 million (EUR 88 million). Further to these calculations, restoration costs related to public water supply and point-of-entry treatment units for 2018 to 2050 was valued at USD 396 million (EUR 288 million). It is unclear in the Sunding study what is covered by these restoration costs. Nevertheless, it is presumed that the Sunding study concern the upgrading of water treatment facilities to remove PFAS. The D4/D5 study using the UK survey, notes the following "Quantification of environmental impacts of regulatory policy changes is difficult. In the case of D4/D5 the benefits of the proposed restriction are estimated by considering society's Willingness to Pay (WTP) for a reduction in potential risks to the aquatic environment. ... A representative sample of the UK population (sample size = 829) stated that they were willing to pay 46 per year per person to reduce the risks associated with the PBT substance - D4, and 40 per year per person to reduce the risks associated with the vPvB substance - D5. The WTP for superior quality personal care products (i.e. those that use D4/D5) was estimated at 5 per year per person. This indicates that respondents value the precautionary benefits of reduced potential risk of accumulation of D4/D5 in the aquatic environment at around seven times the value of the loss of beneficial properties provided by D4 and D5 in personal care products." The application of the results of the survey recognizes that significant uncertainty exists. Rather than seeking to provide a total estimate of damage to compare against any increase in cost of alternatives, the study compares two alternative WTP estimates. The first WTP estimate is linked to avoidance of environmental harm and the second to the useful properties that are linked to the presence of D4/D5 in personal care products, providing a general indication of societal preference. These results are discussed further below, though a full quantification by country is not presented. To make such a quantification based on the Sunding analysis of house prices and fishing trips would require additional information that is not available on the scale of the current assessment (e.g. knowledge of angling behavior in areas where there is PFAS contamination). The results of the D4/D5 assessment are based on a relatively small survey, the authors of the work applied them only in a limited way in development of the restriction dossier, not quantifying a total cost but comparing WTP to avoid damage with WTP for the properties that they provide in the end product. 3.2.4 Value transfer Value transfer, often called benefits transfer in environmental economics, is a practice used to estimate economic values by transferring information available from an already completed study in one context to another, for example by transferring valuation of fishing trips in the USA to a European country.64 The need for value transfer arises because of the limited amount of valuation data available. 64 The methods are described in detail in Navrud S. and Ready R. (eds.) (2007). Environmental Value Transfer: Issues and Methods. Springer. 48 The cost of inaction linked to PFAS exposure At its most basic, values are simply transferred from one location to another, with no adjustment. Such an approach is reasonable where the differences between locations are expected to be small. However, in most cases the process is not as simple, where people in one location may respond differently to those in another location for example relating to differences in income, culture, or availability of alternative resources. The validity of the transfer should therefore be evaluated carefully. Three approaches have been described65: unit value transfer value function transfer; and meta-analysis. Unit value transfer is typically made with adjustment for variation in income: = Where Bp' = adjusted benefit estimate for new location, BS = original estimate from study site, Yp and YS are the income levels at the new location and study site respectively and = elasticity (typically 0.4-1.0). Jacobsen and Hanley66 found that GDP/capita (i.e. societal income) was a better predictor of WTP than respondents' income, which simplifies the calculations. International transfer in this context is carried out using exchange rates adjusted for purchasing power parity (PPP). Purchasing power parities are the rates of currency conversion that equalise the purchasing power of different currencies by eliminating the differences in price levels between countries. The production of PPPs is a multilateral exercise involving the National Statistical Institutes of the participating countries, Eurostat and the Organisation for Economic Co-operation and Development (OECD), and is governed by the PPP Regulation.67 Data are available from OECD.68 In our study setting a question that arises is how reliable extrapolation from the USA to Europe might be, given that a significant part of the data identified for this research comes from North America. One response to this question is that there is likely as much or more variation between European countries, as there is between the USA and Europe. Another response concerns the way that the US data are used, often alongside European data and serving to extend the evidence base for the analysis. Consideration has been given to whether there is any clear disparity in US and European estimates. 65 Navrud S. (2016). Possibilities and challenges in transfer and generalization of monetary estimates for environmental and health benefits of regulating chemicals. (Presentation at OECD Workshop on socioeconomic impact assessment of chemicals management). 66 Jacobsen J B and Hanley N (2009). Are There Income Effects on Global Willingness to Pay for Biodiversity Conservation? Environmental and Resource Economics, Volume 43, Issue 2, pp 137-160. 67 Regulation (EC) No 1445/2007 of the European Parliament and of the Council of 11 December 2007 establishing common rules for the provision of basic information on Purchasing Power Parities and for their calculation and dissemination. 68 OECD Database: Purchasing power parities. Accessed 10.11.2018. The cost of inaction linked to PFAS exposure 49 Value function transfer and meta-analysis add further explanatory variables into the equation to explain how values may vary from site to site. For the present analysis, however, neither technique has proved necessary in a formal sense because of the nature of the data gathered. However, the derivation of ranges and best estimates takes account of some of the principles of value function transfer by recognising non-linearities in costs according to (e.g.) the number of households affected.69 69 Upon request the excelspread sheets used for the monetarisation and valuation in this report can also be provided along with a guidance on how to use the estimation of costs for value transfer. 50 The cost of inaction linked to PFAS exposure 4. The case studies The case studies have multiple roles in this study. One role is to illustrate the impact pathways, i.e., how the production of PFAS and the manufacture of PFAS-containing products, as well as the use and end-of-life disposal of those products, result in emissions to the environment and human exposures. This provides a basis for identifying the impact pathways for similar instances of contamination in other locations. An additional purpose of the case studies is to gather concrete cost data based on actual instances of PFAS exposure and to translate these into costs per incident. They also help to identify the additional information needed to quantity the negative impacts in a way that will enable them to be extrapolated - if possible - to the Nordic region and then to Europe overall. The case studies follow the life cycle of PFAS from their production at chemical production facilities, their application in manufacturing of products, the use phase and the impact pathways for wastewater discharges, and then the end-of-life phase and their disposal in landfills. Incineration is not reviewed per se, as the PFAS emissions from this disposal method (air-borne as well as via the bottom ash) are not yet well studied.70 Several case studies also provided opportunities for getting human epidemiological data on health effects as well as concrete health-related costs. 4.1 Case Study 1: Exposures due to production of PFAS Case Study 1 looks at how the production of PFAS results in emissions to the environment, resulting in human and environmental exposure (impact pathways). 4.1.1 Background and context The production of PFAS can generate extensive emissions. Most PFAS are colorless and odorless71, and initially they were produced without consideration of the impacts of emissions during their production and processing. The US company 3M was the first to manufacture a PFAS commercially - the C8, PFOA. It licensed the technology to another US company, Dupont, which developed the polymer Teflon. In the 1970s and 1980s 3M scientists became concerned because laboratory animals exposed to C8s were developing health problems, including birth defects. They alerted Dupont scientists and Dupont started to take measures to protect 70 Hansson et al. (2016). Sammanstllning av befintlig kunskap om froreningskllor till PFAS-mnen i svensk milj. Report Number C 182. 71 The odor of PFBA has been described as "sharp, similar to butyric acid. The cost of inaction linked to PFAS exposure 51 workers exposed in their jobs. Measures to reduce emissions to the environment were not taken until considerably later. Contamination levels at production sites can be high. A number of lawsuits are underway to prompt action from manufacturers to clean-up contamination. The level of damages and settlements reached depend on a number of factors which include the identified harm to the environment and human health. Some settlements include punitive damages and funds to support the cleanup of the contaminated water. In February 2018, a settlement of USD 850 million (EUR 618 million) was reached between 3M and the state of Minnesota.72 The funds are expected to support the development of alternative water supplies, treat existing water supplies, finance water conservation and efficiency projects and support groundwater recharge projects.73 While it is the largest settlement to date in the United States for a case of PFAS contamination, the amount originally sought by the plaintiff was USD 5 billion.74 The settlement amount diminishes in comparison to the company's market value (about 1%) and the profits made on the PFAS-containing products.75 3M is facing other legal actions: a February 2018 news article reported that 37 cases related to PFC contamination against 3M are underway.76 4.1.2 Cases of contamination Chemours factory complex in Dordrecht, Netherlands The Chemours plant in the Netherlands belonged to the company Dupont until 2015, when Dupont spun off its specialty chemicals division, which included production of fluorochemicals such as the fluoropolymer Teflon. The Dordrecht plant is Chemours' biggest production site in Europe, employing over 550 employees.77 In addition, Chemours has a total of 35 production sites worldwide, 25 in North America, four in Europe, Africa and Middle East, four in Asia and two in South America. In 2012 Dupont had replaced production of C8 with GenX, a perfluoroalkyl ether carboxylate alternative that includes FRD-902, FRD-903 and E1. While GenX seems to be less bioaccumulative than PFOA78, reports filed by Dupont with the USEPA indicated that the replacement chemical may cause some of the same health problems as PFOA.79 Although the Dordrecht facility supposedly stopped PFOA production in 2012, toxicological studies found concentrations of PFOA in grass surrounding the plant that 72 Kary T, 3M Settles Minnesota Lawsuit for $850 million. Bloomberg, 20.02.2018. Accessed 08.13.2018. 73 Marten Law (website). The True Cost of Scotchgard: 3M to Pay Minnesota $850 Million in Perfluorochemical Settlement. Accessed 10.10.2018. 74 Marcotty J, State alleges 3M chemicals caused cancer and infertility, alleges $5 billion in damage. Star Tribune, 21.11.2017. Accessed 10.10.2018. 75 Alder C, 3M settles Minnesota groundwater lawsuit for $850 million. 21.02.2018. Accessed 09.09.2018. 76 Marcotty J, 3M settles groundwater lawsuit for $850 million, Star Tribune, 20.02.2018. Accessed 15.10.2018. 77 Scott A, Dutch Chemical Plant Under Investigation, Chemical and Engineering News, 18.04.2016. Accessed 10.11.2018. 78 Beekman M et al. (2016). Evaluation of substances used in the GenX technology by Chemours, Dordrecht, RIVM Letter. Report 2016-0174. 79 Lerner S (2016). New Teflon Toxin Causes Cancer in Lab Animals, The Intercept, 03.06.2016. Accessed 18.09.2018. 52 The cost of inaction linked to PFAS exposure were inconsistent with that claim.80 Today the facility discharges both FRD-902 and FRD-903 into wastewater, which is pre-treated before being discharged into sewage treatment. FRD-903 is also released into the air via the factory's chimneys.81 In 2016, the Dutch government asked the Dutch National Institute for Public Health and the Environment (RIVM) to evaluate the existing knowledge of the toxicity and health effects of PFOA (C8), which had been released from the Chemours facility from 1970 until 2012.82 RIVM estimated that 750,000 people were exposed to high levels of PFOA due to their residence in cities close to the Dordrecht plant and the Merwede river downstream.83 RIVM also detected PFAS and GenX in vegetable gardens within 1 kilometer of the Dordrecht plants, suggesting pathways into humans.84 While levels of contamination do not exceed advisory thresholds, residents also face exposure from the air directly as well as drinking water. Thus, residents are advised to consume produce from their vegetable gardens in moderation. The Chemours facility used PFOA to produce the fluoropolymer known as Teflon, a production process similar to that carried out at a Chemours (formerly Dupont) plant in West Virginia.85 The West Virginia plant has been the target of legal action since 2001, when residents brought a class action against Dupont concerning their exposure to PFOA. Under a 2004 settlement, Dupont agreed to fund a medical monitoring program for 70,000 persons as well as new water treatment systems. After ten years, the C8 panel of scientists following the medical monitoring program concluded that six illnesses were probably linked to the exposure to PFOA: kidney and testicular cancer, ulcerative colitis, thyroid disease, pregnancy-induced hypertension and high cholesterol. More than 3,500 personal injury claims were pending against Dupont and Chemours when agreement was reached to settle the claims for USD 671 million (EUR 488 million). The Netherlands government requested a medical monitoring study of the health effects resulting from the exposure in Dordrecht, following the approach taken with the Dupont case in West Virginia. The study undertaken by RIVM found the levels of PFOA in Dordrecht's drinking water to have been lower than in the US case. However, concentrations in the atmosphere between 1970 and 2012 had exceeded the legal amount. RIVM's risk assessment concluded that under the most unfavourable scenario for concentration levels, some residents might experience impacts on the liver. However, risks of cancer and to the unborn child were considered to be limited.86 Chemours was required to reduce GenX emissions from 6,400 kg/year to 2,000 kg/year. In September 2018, the company announced it would invest EUR 75 million in 80 No Author, Another Chapter: Chemours releases GenX and PFOAs into waters globally, Encore, 08.08.2018. Accessed 10.10.2018. 81 RIVM (2016). GenX (Website in Dutch). 82 RIVM (2016). Risicoschatting emissie PFOA voor omwonenden (In Dutch). 83 No Author (2018). Another Chapter: Chemours releases GenX and PFOAs into waters globally, Encore, 08.08.2018. Accessed 10.10.2018. 84 RIVM (2018). Risicobeoordeling van GenX en PFOA in moestuingewassen in Dordrecht, Papendrecht en Sliedrecht. (In Dutch with English Summary). 85 Nair A S, DuPont settles lawsuits over leak of chemical used to make Teflon. Reuters Business News, 13.02.2017. Accessed 10.09.2018. 86 RIVM (2018). Risicobeoordeling van GenX en PFOA in moestuingewassen in Dordrecht, Papendrecht en Sliedrecht. (In Dutch with English Summary). The cost of inaction linked to PFAS exposure 53 reducing emissions of GenX and organic fluorinated substances, by installing active carbon filters and other technical solutions which are expected to eventually remove up to 99% of the targeted substances.87 Note that a Chemours (formerly Dupont) facility in Fayetteville, North Carolina is also the target of legal action for polluting a wide geographical area. The plant discharged large amounts of GenX into the Cape Fear River, the major source of drinking water for downstream cities, including Wilmington (110,000 pop.). A lawsuit filed against Dupont and Chemours in February 2018 consolidated three class action suits and is seeking funds for environmental cleanup (cost of water filtration), monitoring, and punitive damages for illness and reduced property values.88 In November 2018, state authorities and Chemours announced they had reached an agreement which will cost the company USD 12 million (EUR 8.7 million) in civil penalties, in addition to USD 1 million (EUR 727,400) in investigation costs. Furthermore, if the agreement is approved by the court in its current state, Chemours has committed to reduce its emissions significantly. If reduction targets are not achieved additional fines will be paid.89 Veneto region, Italy A large-scale contamination of PFAS was discovered in the Veneto Region of Italy in 2013, directly affecting groundwater, surface water, drinking water and land in an area of over 200 square kilometers.90 The contamination was attributed to emissions from a facility operated by the company Miteni since 1964. The chemical company produced several PFAS-containing products such as herbicides and pharmaceuticals. The company reported on their website that production of PFOS and PFOA stopped in 201191, but their product catalog still includes PFHxS and PHxSF.92 The Veneto Region's response to the incident drew on a wide range of stakeholders across sectors. Several monitoring studies were undertaken to gather information on the levels of contamination and their impacts. Because of this contamination, standards were introduced for water, agriculture and food. In April 2018, a legal case was brought against Miteni93 by the Public Prosecutor's Office in Vicenza.94 The contamination included a number of PFAS compounds. Monitoring data collected between 2013 and 2015 identified the following specific compounds: perfluorobutanoic acid (PFBA), perfluorobutane sulfonate (PFBS), perfluorodecanoic 87 Chemours, The Chemours Company Takes Significant Action to Minimize Emissions. Accessed 08.09.2018. 88 Reisch M S, Merged lawsuit filed against DuPont and Chemours in North Carolina, Chemical and Engineering News, 05.02.2018, Accessed 08.09.2018. 89 Wagner A, Chemours to pay $12 Million fine as part of GenX Agreement, StarNews Online, 21.11.2018. Accessed 01.12.2018. 90 WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. 91 As claimed by Miteni Company on their website, available here. 92 This information was obtained by an NGO (IPEN and Alaska Community Action on Toxics (ACAT)) submission of information specified in Annex E of the Stockholm Convention pursuant to Article 8 of the Convention, and confirmed with a product research on the company's website using CAS numbers for PFHxS and PFHxSF. 93 In October 2018, the company filed bankruptcy. 94 No author, La procura di Vicenza indaga sulle sostanze cancerogene nell'acqua, Le Iene, 21.05. 2018, Accessed October 2018 (in Italian). 54 The cost of inaction linked to PFAS exposure acid (PFDA), perfluorododecanoic acid (PFDoDA), perfluoroheptanoic acid (PFHpA), perfluorohexanoic acid (PFHxA), perfluorohexane sulfonate (PFHxS), perfluorononanoic acid (PFNA), perfluoropentanoic acid (PFPeA) and perfluoroundecanoic acid (PFUnDA).95 Another source indicated that PFPeA, PFHxA, PFHpA, PFHxS, PFNA, PFDA, PFUnDA, PFDoDA, PFBA, and PFBS were present in the contaminated waters.96 Residents in the surrounding areas were exposed to the contamination for decades. The highest combined concentration levels of PFAS (estimated to be 1214 ng/l) were found in the municipalities of Brendola, Lonigo and Sarego97 among others.98 Estimates for the number of affected individuals vary from 120,00099 to 350,000 people.100 The Veneto Region Environmental Agency carried out studies identifying two main pathways for the contamination in the area. One pathway was through contaminated wastewater emitted from the chemical factory directly into a creek and the surrounding groundwater. The second pathway was from the wastewater plant to a canal that drained into the surface waters of the Brenta river.101, 102 This river basin, which is characterised by a 100-200 metre deep bed of pebbles and gravel, is highly permeable, increasing the spread of the contamination. The contaminated water was drunk by residents in the area and also made its way into the food chain via contaminated water used for irrigation. Contamination was found in foods such as eggs and fish. Agriculture in the Veneto Region represents about 10% of national production suggesting a high risk for contaminated foods being shipped and consumed in other parts of the country or even in other countries.103 Information on health risks was collected through a biomonitoring study undertaken between July 2015 and April 2016 for different population groups including people working in the agricultural sector and people residing in uncontaminated areas, which served as a control group. Agricultural workers were given greater attention due to their greater risk of exposure to crops and livestock. An analysis found a higher risk of mortality, diabetes, cerebrovascular diseases, myocardial infarction and Alzheimer's disease. Females had an increased morbidity for several conditions. The results are summarized in Table 8. 95 Mastrantonio M et al. (2017). Drinking water contamination from perfluoroalkyl substances (PFAS): an ecological mortality study in the Veneto Region, Italy. The European Journal of Public Health. Feb 1;28(1):180-185. 96 ISS (National Institute of Health). Dept of Environmental Health (no date). Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. 97 Ibid. 98 Regione di Veneto (2018). Piano di Sorveglianza Sanitaria Sulla Popolazione Esposte alle Sostanze Perfluoroalchiliche (in Italian), Rapport no 04, p.10. 99 WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. 100 No author, Another Chapter: Chemours releases GenX and PFOAs into waters globally, Encore, 08.08.2018. Accessed 10.10.2018. 101 WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. 102 ISS (National Institute of Health). Dept of Environmental Health (no date). Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. 103 No author (2018). La procura di Vicenza indaga sulle sostanze cancerogene nell'acqua, Le Iene, 21.05. 2018, Accessed October 2018 (in Italian). The cost of inaction linked to PFAS exposure 55 Table 8: Disease risk ratios - Veneto region Males Females General mortality Diabetes Cerebrovascular disease Myocardial infarction Alzheimer's disease Kidney cancer Breast cancer Parkinson's disease Number of deaths 21,149 292 1,871 1,900 89 Relative risk (95% confidence interval) 1.19 (1.17-1.21) 1.21 (1.06-1.38) 1.34 1.27-1.41) 1.22 (1.16-1.28) 1.33 (1.05-1.70) 1.07 (0.90-1.28) n.a. Number of deaths 20,692 595 2,271 1,458 178 Relative risk (95% confidence interval) 1.21 (1.19-1.23) 1.48 (1.34-1.62) 1.29 (1.23-1.34) 1.24 (1.17-1.32) 1.35 (1.09-1.67) 1.32 1.11 1.35 Source: Mastrantonio M et al. (2017). Drinking water contamination from perfluoroalkyl substances (PFAS): an ecological mortality study in the Veneto Region, Italy. The European Journal of Public Health. Feb 1;28(1):180-185. Monitoring studies also found a higher rate of mortality and disease (liver, bladder and kidney cancer and cirrhosis) among employees at the chemical plant.104 Within three months of the 2013 announcement of PFAS contamination in the Veneto Region, authorities installed activated carbon filters in drinking water treatment plants. The estimated cost of the installation was EUR 2 million, which was paid for by the Veneto Region's government and taxpayers. Activities related to the Health Surveillance Plan cost an additional EUR 4.3 million. These modifications resulted in a substantial decline in the concentration of PFOA in water from about 1475 ng/l to 386 ng/l, and a decline in PFOS from 117 ng/l to 36 ng/l. Maintaining the carbon filters will also incur costs. One source estimated the cost of a changed set of filters at EUR 150,000 while chemical monitoring costs EUR 750,000, leading to a total cost of the filter installation of EUR 900,000 per year. The short-term cost (5 years) was estimated to be EUR 6.5 million while investment in the medium-term to improve water treatment plants was estimated to be EUR 4.2 million. An alternative solution to invest in new water pipelines was estimated to be EUR 61.7 million. Table 9 translates the cost findings to the production cost of drinking water. Table 9: Production cost of drinking water 2014 (EUR cent/m3) Cost without PFAS pollution Cost with PFAS pollution 4.7-8.3 10.0-18.7 2015 (EUR cent/m3) 4.0-8.5 6.6-21.0 Source: ISS (National Institute of Health). Dept of Environmental Health (no date). Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. 104 WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. 56 The cost of inaction linked to PFAS exposure Between 2013 and 2014, the chemicals company also installed activated carbon filters for wastewater treatment.105 Fluorochemical production sites - data limitations and assumptions As the case studies indicate, fluorochemical production plants have historically been major sources of PFAS exposure. One of the challenges of this study was to estimate the current number of fluorochemical production sites in the Nordic countries and in Europe. The constant changes within the fluorochemicals/fluoropolymers industry made it difficult to use data sources from the past that identified sites by company. For example, Elf Atochem had merged with Total/Fina to become Alofina which later became, Arkema. Similarly, ICI Fluorochemicals was sold to INEOS, which in turn sold that part of its business to Mexichem. In addition, production is shifting from Europe and North America to Asia, particularly China, and other regions. Among the major producers are companies such as Daikin, Asahi Glass, Honeywell, Mitsui Chemicals, Gujaratfluorochemicals, and Dongyue Chemicals. BASF, Solvay, and Dupont are also global players. The US-based FluoroCouncil represents a number of PFAS producers internationally. Members include Archroma, Arkema France, AGC Inc., Chemours, Daikin Industries, and Solvay Specialty Polymers. Dynax Corporation and Tyco Fire Products are associate members. Regulatory pressure has also contributed to changes in the industry. Production of the long-chain PFOA and PFOS was phased out years ago by companies like 3M, DowDupont, and BASF, which for the most part switched to production of short-chain PFAS. However, company websites seldom provide information on which fluorochemicals are produced and at which sites. The list presented in Table 10 is the result of the search efforts carried out for this study. It assumes that production of fluorochemicals and/or fluoropolymers is taking place in 2018 at the following sites within Europe: 105 ISS (National Institute of Health). Dept of Environmental Health (no date). Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. The cost of inaction linked to PFAS exposure 57 Table 10: List of manufacturers of fluorochemicals and/or fluoropolymers Country Company and site of plant What is being produced Belgium France Germany Italy Netherlands United Kingdom 3M (Zwijndorf) Arkema (Pierre-Bnite) Solvay Solexis (Tavaux) Daikin Chemical France S.A.S.1 Dyneon (Gendorf) BASF (Ludwigshafen) Solvay Solexis (Spinetta-Argeno Heroflon S.p.A. (Collebeato) Miteni (Trissino) 2 Chemours (Dordrecht) Daikin Chemical Netherlands (Oss) - Pre-compounding of fluoroelastomers AGC (Blackpool) Fluorochemicals Fluoropolymers (PVDF) Fluoropolymers (PVDF) Fluorochemicals Fluorochemicals, fluoropolymers (PTFE, FEP, PFA, THV) n.a. Fluoropolymers ) - PTFE, MFA Fluoropolymers (PTFE compounds and micropowders) Fluorinated intermediates; performance fluorinated products Fluoropolymers (PTFE, FEP) Fluorochemicals Fluoropolymers - PTFE, PFA Source: 1) Daikin Europe, manufacturing of fluoroelastomer base polymer and polymer processing aids. 2) Miteni files bankcruptcy in October 2018. Based on this list, it is further assumed that the number of PFAS production sites in Europe is between 12 and 20 plants. However, in this study the authors have not been able to identify any PFAS production facilities in the Nordic countries. 4.2 Case Study 2: Exposures due to manufacture and commercial use of PFAS-containing products 4.2.1 Background and context The many facilities where PFAS are used in the manufacture of consumer goods and other products also constitute major sources of PFAS emissions to the environment. A 3M study from 2000 estimated that 15% of all indirect emissions of POSF (perfluorooctane sulfonyl fluoride, compound that is used for producing PFOS) occurred during manufacturing from secondary applications.106 In addition, the use of PFAS as surfactants and coatings in providing commercial services, such as professional cleaning, has also been linked to hotspot contamination. A mapping of PFAS pathways in Sweden concluded that quantifying the emissions released from the many uses of PFAS in products might not be possible due to lack of data and the "diffuse character" of certain areas of usage.107 To increase information on how and where PFAS are used, the Swedish Chemicals Agency has recently introduced 106 Alexandre G. Paul et al. (2008). A First Global Production, Emission, And Environmental Inventory For Perfluorooctane Sulfonate. Environmental Science & Technology 2009 43 (2), 386-392. 107 Hansson et al. (2016). Sammanstllning av befintlig kunskap om froreningskllor till PFAS-mnen i svensk milj. Report Number C 182. 58 The cost of inaction linked to PFAS exposure a requirement for companies to report all products containing added PFAS compounds and their use to the Swedish "Products Register"108, starting from 1 January 2019. This case study identifies several industrial activities that have the potential to release PFAS to the environment, either directly or indirectly, e.g., through discharges to wastewater treatment works on site or off-site via sewerage systems. It provides brief descriptions of cases where contamination has been identified, leading to costs for remediation. 4.2.2 Cases of contamination Textile and leather manufacturing The textile industry is one of the most extensive users of PFAS. The water repellency and stain resistance qualities of PFAS have led to widespread PFAS treatment of items intended for use outdoors such as raincoats, snowsuits for children, ski jackets, shoes and umbrellas as well as outdoor equipment such as tents, awnings and sails. Other items frequently treated with PFAS include carpets, upholstery, and leather products. Textile applications can account for an estimated 35% of the demand for fluorotelomers. This demand is projected to grow by 14% by 2020.109 Awareness within the textile industry concerning the health and ennvironmental impacts of PFAS is growing and alternative non-florinated methods of treating fabrics and leather for water repellency have been analysed. Though the case study of PFAS contamination in the Veneto region above focused on the facility producing fluorochemicals operated by Miteni, sources of the contamination may also include hundreds of smaller textile and leather companies. In fact, the Veneto Region is known for its model of "industrial clusters" where multiple companies producing the same good are located in the same area. As shown in Figure 3 below, several industrial clusters producing goods that could contain PFAS were present in the contaminated area. 108 Swedish Chemicals Agency (2018). (News) The Swedish Chemicals Agency is introducing a requirement to report PFAS to the Products Register. Accessed September 2018. 109 Press release (No date) Fluorotelomers Market to Reach USD 539.3 Million Worldwide by 2020, Digital Journal. Accessed 10.11.2018. The cost of inaction linked to PFAS exposure 59 Figure 3: Industrial clusters that could contribute to PFAS contamination pathways Source: ISS (National Institute of Health). Dept of Environmental Health (no date). Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. Miteni's website110 includes a statement from April 2016 expressing the opinion that the large area contaminated by PFAS "must necessarily refer to the system of consortium discharges to which hundreds of local companies are connected. Miteni has not been producing PFOS and PFOA since 2011, and before, the wastewater from the facility was sent to external treatment systems. PFOS and PFOA are still used today by over two hundred industries in the tanning and manufacturing industries in the area that buy them on the foreign market, companies that are connected to the same consortium discharges to which Miteni is connected." The proportion of textile and leather manufacturers in Europe using PFAS to treat their products is not known, but even a small percentage such as 3% could be significant. Statistics from Eurostat111 indicate 262 textile manufacturers in the EEA with more than 250 employees in 2015, and another 61,685 textile manufacturers with less than 250 employees, i.e., SMEs.The country with the highest number of textile manufacturers is Italy (13,930 textile manufacturers). Manufacturers of leather and leather goods in 2015 included 159 companies with more than 250 employees, and another 37,120 companies with less than 250 employees. Case Study 4 below looks at carpets, a textile-related consumer product, while Case Study 5 includes a description of historic pollution in the State of Michigan involving a former leather tannery. 110 Miteni Company Website. 111 Eurostat Annual enterprise statistics by size class for special aggregates of activities (NACE Rev. 2) [sbs_sc_sca_r2]. 60 The cost of inaction linked to PFAS exposure Metal plating, including chromium plating The use of PFOSs as a wetting agent/fume suppressant in chromium plating was first reported in 1954112, and additional types of PFAS have been developed for such use since then.113 PFAS are also used to improve the stability of the baths used in electroplating of copper, nickel and tin and the overall performance of the process. In addition, PFAS are used to treat metal surfaces to prevent corrosion, reduce mechanical wear, or enhance aesthetic appearance, as well as to promote flow of metal coatings and prevent cracks during drying.114 The chrome plating industry is estimated to use around 32 to 40.7 tonnes of pure PFOS globally. The estimations are likely to cover different applications within metal plating, and not only chromium plating.115 For Denmark, this number was estimated to be 10 kg/year in 2009 with 28 kg/year as purchased quantity.116 In Sweden, the figure (from 2013) was 180 kg/year for PFOS.117 Again, the proportion of metal plating facilities in Europe using PFAS is not known. Statistics from Eurostat118 on the number of establishments of treatment and coating of metals and machining in the EEA indicate 163 establishments with more than 250 employees in 2015, and another 151,455 establishments with less than 250 employees, i.e., SMEs. Paper and paper products Paper mills and paper products industry using PFAS to produce waterproof and greaseproof paper products may be a significant source of PFAS contamination released to water and air as well as to soil.119 For example, a mapping of PFAS pathways to Norway's Tyrifjorden identified paper mills as among the main sources of the fjord's PFAS contamination. As per Case Study 4.5.2.1 below, paper mill waste has also been implicated in PFAS-contaminated substrate sold as compost. The PFAS are typically added to wet wood fibers that are subsequently made into paper. A US NGO obtained notifications to the US Food and Drug Administration concerning PFAS for use in food contact materials.120 The FDA notifications estimated releases to the environment of the PFAS based on a typical paper mill producing 825 tonnes of PFAS-coated paper per day and discharging 26 million gallons (around 99,000 m3) of water per day. One notification from the company Chemours estimated wastewater discharges of 95 pounds/day (43 kg/day-15,965 kg/y) in discharges contain- 112 German Environment Agency (2017). Use of PFOS in chromium plating - Characterisation of closed-loop systems, use of alternative substances. Report No. (UBA-FB) 002369/ENG. 113 Danish Environmental Protection Agency (2011). Substitutions of PFOS for use in non-decorative hard chrome plating. 114 Kissa E (2001). Fluorinated Surfactants and Repellents. Surfactant Science Series. 97. Marcel Dekker, New York 115Danish Environmental Protection Agency (2011). Substitutions of PFOS for use in non-decorative hard chrome plating. 116 Ibid. 117 Hansson et al. (2016). Sammanstllning av befintlig kunskap om froreningskllor till PFAS-mnen i svensk milj. Report Number C 182. 118 Eurostat Annual enterprise statistics by size class for special aggregates of activities (NACE Rev. 2) [sbs_sc_sca_r2]. 119 Neltner T and Maffini M, Paper mills as a significant source of PFAS contamination, but who's watching? EDF, 21.05.2018 Accessed 10.11.2018. 120 Ibid. The cost of inaction linked to PFAS exposure 61 ing 43,000 ppt. Another notification for the same PFAS estimated higher concentrations in the paper, resulting in 183 pounds/day (83 kg/day-30,295 kg/year) and wastewater discharges containing concentrations of 83,000 ppt. Paper and cardboard treated with PFAS are frequently used in products such as plates, popcorn bags, pizza boxes, and food containers and wraps (see Case Study 4 on food contact materials). Non-food applications may include folding cartons and masking papers.121 Statistics from Eurostat122 on the number of manufacturers of paper and paper products in the EEA provide totals of 488 manufacturers employing more than 250 employees in 2015, and another 19,477 establishments with less than 250 employees, i.e., SMEs. It is not known how many of these manufacturers use PFAS to treat their products. Paints and varnishes PFAS have long been used in coating, paint, and varnish to reduce surface tension for substrate wetting and levelling, as dispersing agents, and for improving gloss and antistatic properties. The protective properties of anticorrosive paints can be enhanced by perfluorinated urethanes. PFAS can also be used as aids in pigment grinding and to address pigment flotation problems.123 Eurostat statistics124 indicate that 104 manufacturers of paints and varnishes in the EEA employ more than 250 employees in 2015, and another 4,027 establishments employ less than 250 employees. Again, it is not known how many of these use PFAS surfactants in their products. Cleaning products The surfactant properties of PFAS have made them useful in industrial and household cleaning products. They have been used to lower surface tension and improve wetting and rinse-off in products such as carpet spot cleaners, alkaline cleaners, denture cleaners and shampoos, floor polish, and dishwashing liquids. They are sometimes added to cleaners containing strong acids and bases, such as those for cleaning concrete, masonry, and metal surfaces (e.g. airplanes). They may also be used in nonaqueous cleaning agents to aid in removal of adhesives and in dry cleaning of textiles or in the cleaning of metal surfaces. PFAS can also enhance cleaning formulations for removal of calcium sulfate scale from reverse osmosis membranes. The relevant Eurostat category here is "Manufacturers of soap and detergents, cleaning and polishing preparations, perfumes and toilet preparations".125 Under this category, some 178 manufacturers in the EEA employ more than 250 employees in 2015, and another 9,402 establishments employ less than 250 employees. Again, it is not known how many of these use PFAS surfactants in their products. 121 Kissa E (2001). Fluorinated Surfactants and Repellents. Surfactant Science Series. 97. Marcel Dekker, New York. 122 Eurostat Annual enterprise statistics by size class for special aggregates of activities (NACE Rev. 2) [sbs_sc_sca_r2]. 123 Ibid. 124 Ibid. 125 Ibid. 62 The cost of inaction linked to PFAS exposure Plastics, resins, and rubbers PFAS are used to manufacture certain plastics or applied plastics such as polytetrafluoroethylene (PTFE) and polyvinylidene fluoride (PVDF). PTFE is best known by the brand name Teflon (Dupont, now Chemours). It has hundreds of uses in consumer and industrial products such as textiles, medical equipment, cookware, and so on. PVDF is used in a range of industrial applications such as automotive fuel hoses, electrical cable insulation and jacketing, high purity piping, and semiconductor piping.126 They are also used as mold-release agents for thermoplastics, polypropylene, and epoxy resins, polyurethane elastomer foam molding, in formulations for antiblocking agents for vulcanized and unvulcanized rubbers, in silicone rubber sealants for soil resistance, and to improve wetting of fibers or fillers in composite resins.127 The closest Eurostat category here is "Manufacturers of basic chemicals, fertilisers and nitrogen compounds, plastics and synthetic rubber in primary forms".128 Under this category, 340129 manufacturers in the EEA had more than 250 employees in 2015, and another 8,650 establishments employ less than 250 employees. Again, it is not known how many of these are manufacturers using fluorochemicals or fluoropolymers. Car washes The surfactant properties of PFAS make them useful for a wide range of industrial cleaning products and surface treatments. For example, they are used in car wash products and automobile waxes, which makes car washes potential sources of PFAS contamination. In 2018, a car wash facility in the US state of New Hampshire was cited as one of the sources of PFAS contamination in wells serving several nearby towns.130 Investigators tested wells on the car wash property and found levels for PFAS higher than expected - up to 158.8 ppt, compared to the USEPA lifetime advisory level of 70 ppt. The facility will be required to take measures to prevent the contamination from continuing. According to the International Car Wash Association 79,000 car wash facilities are operating in Europe. These are likely to be SMEs employing less than 250 workers.131 It is not known how many of these that use products containing PFAS. Manufacturers of PFAS-treated products - data limitations and assumptions Table 11 present the number of manufacturers carrying out activities that may involve the use of PFAS. These statistics are taken from Eurostat and, in the case of car washes, from the industry trade association. The numbers in column 1 represent small-sized manufacturers with less than 250 employees, while column 2 represent medium-sizedmanufacturers with more than 250 employees. 126 The European Commission-DG Enterprise and Industry (2010). Analysis of the risks arising from the industrial use of Perfluorooctanoic acid (PROA) and Ammonium Perfluorooctanoate (APFO) and from their use in consumer articles. 127 Kissa E (2001). Fluorinated Surfactants and Repellents. Surfactant Science Series. 97. Marcel Dekker, New York. 128 Eurostat Annual enterprise statistics by size class for special aggregates of activities (NACE Rev. 2) [sbs_sc_sca_r2]. 129 Figures for CZ, IE, EL and SK are missing for entreprises with 250 or more employees. 130 Sullivan M, Car wash cites for PFAS pollution, Sea Coast Online, June 4 2018. Accessed 19 August 2018. 131 International Car Wash Association. Industry Information. The cost of inaction linked to PFAS exposure 63 Table 11: Number of manufacturers carrying out activities that may have involved use of PFAS Industrial activity Manufacturers <250 employees Manufacturers >250 employees Textiles Leather Carpets Paper Paints and varnishes Cleaning products Metal treatments Car washes Plastic, resins, rubbers Totals 3% of total 10% of total 61,685 37,120 no data 19,477 4,027 151,455 79,000 352,764 10,583 35,276 262 159 488 104 178 163 340 1694 51 169 It is not known how many of the products manufactured by these companies that have been produced or treated with PFAS, or how many commercial service industries that are using products containing PFAS. In the absence of more data, it was assumed that between 3% and 10% of the manufacturers in these industries have used or are using PFAS in the manufacturing process or in the commercial service provided. These activities would have the potential for releasing PFAS to air, water and soil during the manufacturing process as well as afterwards, in the form of industrial waste. Note that these numbers do not include manufacturers from other industrial activities known to use PFAS, such as makers of electronic chips, cosmetics and personal care products, photography films and mineral extraction. Note also that non-fluorinated alternatives are available for some of these applications, e.g., the consumer textile industry. 4.3 Case Study 3: Contamination from use of aqueous film-forming foams 4.3.1 Background and context Aqueous film-forming foams (AFFFs), a specific sub-type of firefighting foam132, are one of the many industrial products that contain PFAS. PFAS-containing AFFFs were found to be particularly useful in extinguishing petroleum-based fires (also known as class B liquid fires).133 The surfactant properties of the PFAS serve to form a coating or blanket that deprives the fire of oxygen until the flames die out. Because of their 132 Norden (2013). Per- and polyfluorinated substances in the Nordic Countries: Use, occurrence and toxicology. TemaNord 2013:542. 133 Swedish Environmental Institute (2015). Risks and Effects of the dispersion of PFAS on Aquatic, Terrestrial and Human populations in the vicinity of International Airports: Final Report of the RE-PATH project. p.17. 64 The cost of inaction linked to PFAS exposure effectiveness, they have been widely used around the world since the 1960s134 both at airports and at fire training facilities. The biggest concern related to the use of AFFFs is the contamination of drinking water sources.135 This concern is also related to long-lasting impacts of these substances in the water. For instance, PFOAs are said to have a hydrolytic half-life of more than 97 years.136 Contamination of the environment due to AFFFs might take place in different ways in specific locations due to local, complex interactions within the ecosystems. Various exposure pathways have been identified.137 The hazardous substances in the AFFFs usually find their way to surface waters through direct runoff, in ground waters through infiltration and in soil through soil diffusion or dispersion. Drinking water can be contaminated as a result of ground water and surface water contamination.138 Groundwater and surface water contamination can also lead to contamination of agricultural produce through uptake into biota. Contaminated surface waters and biota (fish) can create further health risks for humans and other animals.139 A Swedish study ("RE-PATH") investigated contamination around two airports where AFFFs had been used for many years.140 The results clearly indicated increased levels of PFAS in the surrounding soil, surface/ground water and biota - up to 100 times higher than in control areas. Water pollution (whether it is groundwater or surface waters) related to the use of AFFFs following a fire, accidental leaks or fire-fighting trainings is considered to be severe.141 Cases of groundwater, soil and surface water contamination have been documented near airports, military bases and fire drill sites in Germany, the Scandinavian countries142, the UK and others. One report suggests that many classes of PFAS are observed in groundwater, essentially every AFFF impacted site investigated to date.143 As the environmental and health risks of PFAS became apparent in the 1980s, the role of AFFFs in PFAS-related environmental contamination started to draw attention. In Europe, AFFFs containing PFOS were banned in 2006 with a complete phase-out in 134 Krrman A et al. (2011). Environmental levels and distribution of structural isomers of perfluoroalkyl acids after aqueous fire-fighting foam (AFFF) contamination. Environ. Chem., 8, 372-380. 135 Field, J et al. (2017). FAQs Regarding PFAS Associated with AFFF Use at U.S. Military Sites, Report for Environmental Security Technology Certification Program (ESTCP). 136 Franko et al. (2012). Dermal Penetratoi potential of Perfluorootanoic Acid (PFOA) in Human and Mouse Skin, Journal of Toxicology and Environmental Health, Part A. 75:50-62. 137 The National Academies of Science Engineering and Medicine (2017). Use and Potential Impacts of AFFF Containing PFAS at Airports. 138 Interstate Technology Regulatory Council (no date). Environmental Fate and Transport for Per- and Polyfluoroalkly Substances. 139 The National Academies of Science, Engineering, Medicine (2017). Use and Potential Impacts of AFFF Containing PFAS at Airports. 140 Swedish Environmental Protection Agency (2015). Risks and Effects of the dispersion of PFAS on Aquatic, Terrestrial and Human populations in the vicinity of International Airports: Final Report of the RE-PATH project. Report number: B 2232. 141 Eschauzier C et al. (2012). Polyfluorinated Chemicals in European Surface Waters, Ground and Drinking Waters. In: Knepper T., Lange F. (eds) Polyfluorinated Chemicals and Transformation Products. The Handbook of Environmental Chemistry, vol 17. Springer, Berlin, Heidelberg. 142Danish Environmental Protection Agency (2014). Screeningsundersgelse af udvalgte PFASforbindelser som jord- og grundvandsforurening I forbindelse med punktkilder (In Danish). 143 Field J et al. (Report for Environmental Security Technology Certification Program-ESTCP) (2017). FAQs Regarding PFASs Associated with AFFF Use at U.S. Military Sites. p:8. The cost of inaction linked to PFAS exposure 65 2011.144 Starting from 2020, the use of AFFFs containing more than 25 ppb of PFOA or its salts as well as those containing more than 1000 ppb of one or a combination of PFOA related substances will be restricted within the EU.145 The restrictions will not apply to AFFFs placed on the market before July 2020.146 The new generation of AFFFs may still contain shorter chain PFAS.147 However, only limited data is available on the chemical composition of these newer generation AFFFs, previous studies found high concentrations of PFHxA, C6 and 6:2 FTS in the products.148 Little information is also available for the impact of these fluorinated replacement substances on the environment and human health. Furthermore, because the older generation PFAS (such as PFOS) are highly toxic and persistent; the impacts of the historical releases continue to present serious risk to human health and to the environment. On the positive side, highperformance non-fluorinated AFFFs have been developed and are now on the market.149 4.3.2 Cases of contamination Kallinge-Ronneby Military and Civilian Airbase Between 1980 and 2003, AFFFs containing PFOS were used in Sweden. In 2003, these were replaced a new type containing other PFAS but old stocks containing PFOS continued to be used.150 Swedish airports and Swedish armed forces started using a fluorine-free alternative to fluorine-based fire-fighting foams as of June 2011.151 During a 2013 groundwater quality survey, high concentrations of PFAS were detected in the Bredkra delta (Ronneby). The testing was expanded to a larger area, confirming the PFAS contamination in the outgoing water from one of the two municipal waterworks, Brantafors, which supplied water to around 5,000 people.152 Table 12 presents the levels for some of the PFAS compounds detected in Brantafors compared to Krragrden, the second source of drinking water in the municipality. It can be seen that PFHxS, PFOA and PFOS were sometimes 100-300 144 Swedish Chemicals Agency (2015). Survey of Fire Fighting Foam. 145 Commission Regulation (EU) 2017/1000 of 13 June 2017 amending Annex XVII to Regulation (EC) No 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) as regards perfluorooctanoic acid (PFOA), its salts and PFOA-related substances. 146 Ibid. 147 Swedish Environmental Protection Agency (2015). Risks and Effects of the dispersion of PFAS on Aquatic, Terrestrial and Human populations in the vicinity of International Airports: Final Report of the RE-PATH project. Report number: B 2232. 148Swedish Chemicals Agency (2014). Chemical Analysis of Selected Fire-fighting Foams on the Swedish Market. 149 IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 150 Swedish Environmental Protection Agency (2015). Risks and Effects of the dispersion of PFAS on Aquatic, Terrestrial and Human populations in the vicinity of International Airports: Final Report of the RE-PATH project. Report number: B 2232. 151 Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. Report 7/15. 152 Ronneby Municipality Website, page dedicated to PFAS contamination, available here. 66 The cost of inaction linked to PFAS exposure times higher in the contaminated water source. Given that firefighting foams containing PFOS were phased out between 2003 and 2008153, these highly elevated levels underline the persistent nature of PFAS.154 Table 12: PFAS levels (ng/l) in outgoing drinking water from the waterworks in Ronneby, Sweden on Dec 10, 2013 Brantafors Krragrden PFPeA (Perfluoropentanoic acid) 38 10 PFHxA (Perfluorohexanoic acid) 320 3.6 PFHpA (Perfluoroheptanoic acid) 32 1.4 PFOA (Perfluorooctanoic acid) 100 1 PFBS (Perfluorobutane sulfonic acid) 130 <2.6 PFHxS (Perfluorohexane sulfonic acid) 1700 4.6 PFOS (Perfluorooctane sulfonic acid) 8000 27 PFHpS (Perfluoroheptane sulfonic acid) 60 <1 The Swedish National Food Agency sets the recommended action level of 90 ng/l for a combined sum for 11 different PFAS, highlighted in red in the table. Source: University of Goteborg, The Sahlgrenska Academy of Instutite of Medicine (Li Ying et al.), Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water, 2017. The source of the contamination was identified as the fire drill site located in the nearby military airport.155 The contamination is estimated to have started in the mid-1980s.156 The quantity released into the environment is unknown. The PFAS leaked from the site to the surrounding area into soil, eventually reaching the ground waters. The contaminated water source was equipped with a carbon filter at the time and according to the authorities, this filtering system might have reduced contamination of the drinking water, until it became saturated.157 Brantafors waterworks was closed right after the contamination was detected, and reopened a year later in 2014 after being equipped with charcoal filters and using only two of its four abstraction points (the remaining two having too high levels of PFAS). The water was monitored until October 2014, when the PFAS levels showed an upgoing trend which led to the re-closing of the waterwork, despite levels never exceeding the Swedish recommended action level of 90 ng/l. To secure drinking water supply, new pipes were built to provide Brantafors with uncontaminated water from the area Karlsns. Biomonitoring between 2014 and 2016 resulted in testing of 3418 persons to determine exposure to PFAS from the drinking water. In addition, a smaller sub-sample of 106 individuals was created for a panel study to estimate half-life of the substances. These individuals gave regular blood samples and the monitoring will continue in the 153 Banzhaf S et al. (2016). A review of contamination of surface-, ground-, and drinking water in Sweden by perfluoroalkyl and polyfluoroalkyl substances (PFASs). The Royal Swedish Academy of Sciences. p.337. 154 Li Ying et al. (2017). Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. 155 Swedish Chemicals Agency (2015). Survey of Fire Fighting Foam. 156 Li Ying et al. (2017). Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. 157 Ronneby Municipality Website, page dedicated to PFAS contamination, available here. The cost of inaction linked to PFAS exposure 67 coming years.158 The concentrations for the Ronneby main group and for the sub-group were much higher when compared to the reference group, a sample of 242 people living in a nearby community not affected by the contamination, see Table 13. Table 13: Median levels for PFAS concentrations in ng/ml in serum samples, tested 6th months after end of exposure Contaminants Reference Population (n=242) Main Ronneby Group (n=3418) Panel Study Group (n= 106) PFHxS 0.84 PFOA 1.59 PFOS 4.21 152 277 10.4 17.5 176 345 Source: University of Goteborg, The Sahlgrenska Academy of Instutite of Medicine (Li Ying et al.), Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water, 2017 Another pilot study was conducted among school children comparing those who had been drinking the contaminated water and those who had not. The children exposed to contamination were found to have 20-50 times higher levels of different PFAS substances, particularly PFOS, PFOA and PFHxS.159 Ronneby and other incidents led the Swedish authorities and other responsible parties (like the Swedish Air forces) to conduct national scale monitoring of drinking waters as well as further investigations in known contaminated areas.160 The Swedish Chemical Agency (KEMI) and the National Food Agency set up a national PFAS network which brings together a wide range of stakeholders to advance existing knowledge on the issue.161 In one of the few examples of national monitoring of PFAS in the environment, around 6,000 measurements of surface and ground water were compiled (existing data or new measurements) across the county by the Swedish Environmental Protection Agency. All water supplies with contamination levels that exceeded the safety level of 90 ng/l were found to be located close to an individual fire training site or to one located within an airport. The use of fire extinguishing foams was identified as the largest direct point source of contamination.162 Between 2013 and 2015, new water pipe connections were built between uncontaminated wells in Karlsns and Brantafors.163 The cost of changing the water supply from Brantafors to Karlsns is roughly estimated to have cost Ronneby municipality 158 Li Ying et al. (2017). Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. 159 Jakobsson K et al. (2014). Exponering fr perfluorerade mnen (PFAS) i dricksvatten i Ronneby kommun (In Swedish). Rapport 8:2014. 160 Banzhaf S et al. (2016). A review of contamination of surface-, ground-, and drinking water in Sweden by perfluoroalkyl and polyfluoroalkyl substances (PFASs), The Royal Swedish Academy of Sciences. 161 Sahlin S (2017). PFAS in the Baltic Sea Region. 162 Swedish Environmental Protection Agency (2016). Hgfluorerade mnen (PFAS) och bekmpningsmedel en sammantagen bild av frekomsten i miljn. (report in Swedish with English Summary). 163 Ronneby Municipality (No date). Frgor och svar om PFAS. Accessed August 2018. 68 The cost of inaction linked to PFAS exposure SEK 60 million (incl. VAT) (EUR 5.8 million).164 The additional annual cost for increased monitoring is calculated to around SEK 50,000 (incl VAT) (EUR 4,800). According to the latest information available, the investigations are in progress as to how to clean up the contaminated soil in Ronneby. The Swedish Armed Forces is leading the process. Significant water resources remain unusable for an unforeseeable future due to PFAS contamination. The loss of these valuable resources has not been monetised, but should nevertheless be taken into account. Jersey Civilian Airport, Channel Islands The Jersey Airport case is one of the earliest and well documented examples in Europe of contamination of groundwater and surrounding areas due to AFFFs. In 1991, the fire training site started using AFFFs to meet the requirements of UK Airport Fire Services.165 They were regularly using AFFFs until 1993, when foaming water started to emerge from a land drain excavated near the training site.166 A brown coloration and substantial foaming was also identified in a farm's private water supply that was found to be contaminated.167 According to the accounts of the airport, 78 properties were within the plume area. Groundwater in 36 of these properties tested positive for PFOS.168 Although at some of the sites, concentrations of PFOS have shown signs of decline, they have remained at high levels for seven years in private wells.169 For instance, one private well continued to show high levels of PFOS since the first publicly available recording began in 1999 until 2006, between 2.7 g/l (2700 ng/l) and 9.5 g/l (9500 ng/l). In another area where samples were taken, one property had levels as high as 98 g/l (98,000 ng/l).170 As regards groundwater, the level detected in a borehole was 96 g/l (96000 ng/l).171 The fire-training site was identified as the origin of the contamination. The foam used at the site during training exercises was discharged regularly without monitoring172, dissolving into the ground and rainwaters.173 Contamination subsequently found its way into the St.Ouen's aquifer and the beach of St.Ouen's Bay.174 A monitoring program was put in place to regularly test the ground waters for contamination levels, starting from 1994. The impact on agricultural products (potatoes 164 Schyberg, I. (2018). Ronneby Milj och Teknik AB, personal communication. 165 States of Jersey, States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. 166Foundation for Water Research (2008). Survey of the prevalence of perfluorooctane sulphonate (pfos), perfluorooctanoic acid (pfoa) and related compounds in drinking water and their sources. 167 States of Jersey, States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. 168 Jersey Airport Website dedicated to the PFOS, available here: http://www.jerseyairport.com/PFOS/Pages/Questionsand-Answers.aspx 169Foundation for Water Research (2008). Survey of the prevalence of perfluorooctane sulphonate (pfos), perfluorooctanoic acid (pfoa) and related compounds in drinking water and their sources. 170 Ibid. 171 Ibid. 172 Robins N S (British Geological Survey)(2000). The water resources of Jersey: an overview. 173 States of Jersey, States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. 174 Foundation for Water Research (2008). Survey of the prevalence of perfluorooctane sulphonate (pfos), perfluorooctanoic acid (pfoa) and related compounds in drinking water and their sources. The cost of inaction linked to PFAS exposure 69 and cauliflowers etc.) was also analysed. At the time, the monitoring committee could not find any laboratory to test the relevant substances in Europe and the samples had to be sent to the US. In 1999 the authorities started to conduct their own analysis.175 The initial response was to inform the residents using the water supplies of the possibility of contamination and to provide the particular farm mentioned above with a new borehole. Domestic water supplies were tested and free bottled water was offered in cases of contamination.176 In the aftermath of the contamination, Jersey Airport has funded remediation and other related costs.177 To ensure safe drinking water to the residents in the affected area, 67 among 78 properties in the plume area have been connected to the main water supply.178 Long-term remediation works began in 2002 and finished in 2004 with the aim of addressing the current contamination and preventing future problems.179 The total cost was GBP 7.4 million (EUR 10.6 million). Table 14 presents the main breakdowns. Table 14: Details of clean-up costs from the Jersey Civilian Airport Expenditure Type Investigation (incl. installing and monitoring boreholes, ongoing sampling and analysis) Connection to water supplies Payments of Water Rates to Jersey Water Remedial Works associated to Old Fire Training Ground Professional Fees Finance Costs Jersey Water Mains Connection Costs Capital Works - Fire Training Ground Total Total - Euros* 1,427,167 208,237 523,386 446,739 791,778 122,282 808,303 6,292,376 10,620,271 Note The original figures were in British Pounds. They were converted to Euros using the rates from 19 October 2004, the day the Settlement Deed was approved between the parties involved. Source: The Jersey Airport Website Schiphol Airport, The Netherlands In July 2008, an error in the sprinkler-system at a KLM hangar at Schiphol-Ost released 10,000 liters of aqueous fire-fighting foam, containing 143 kg of PFOS, into the surrounding environment.180 This fed into a larger reserve of waste water (100 million liters) kept in five reserve reservoirs, several of which leaked and caused substantial contamination of the soil and surface water.181 A later study found the water resources 175 Ibid. 176 States of Jersey, States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. 177 According to the Jersey Airport Website, accessible here 178 Jersey Airport Website dedicated to the PFOS, available here 179 States of Jersey, States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. 180 Brandveilig, Zorgen om verontreiniging door incident met sprinklerinstallatie 07.09.2011 and No Author, Bodemvervuiling bij Schiphol snel aanpakken, De Volkskrant, 12 October 2011. Accessed 09.11.2018. 181 The Government of Netherlands (2018). Beantwoording Kamervragen over berichten over met PFOS en PFOA verontreinigde grond bij Schiphol. (In Dutch). 70 The cost of inaction linked to PFAS exposure to contain over 12 times the average amount of PFOS otherwise found in several reference sites in the Netherlands.182 Residents were warned in 2008 not to swim in or consume fish from the nearby "Ringvaart" canal until the contaminated waste-water could be drained from the overfull reservoirs.183 Contaminated soil from this incident also resulted in delays of over a year to a project to build a new bus lane in Schiphol-Oost in 2017. Over 50,000 m3 of the soil dug up was found to be contaminated and thus difficult to dispose.184 The cost of the remediation is estimated at EUR 30-40 million.185 Information on AFFFs releases - data limitations and assumptions According to the latest Eurostat data from 2015, Europe has a total of 455 civilian airports (28 MS and Iceland, Norway and Switzerland) with a passenger capacity of 15,000 and more per year), among those 318 are considered main airports with 150,000 or more passengers by year.186 On-line country by country research found an estimated 239 military airfields in the EEA countries and Switzerland.187 The numbers are likely to be an underestimate, because they do not contain non-active military bases, which may be historical sources of AFFF contamination. Also, the US military uses firefighting foams during training and emergencies, as well as in automated fire suppression systems.188 Under US military specifications, AFFFs purchased for use at US military sites must be based on PFAS chemistry to conform to military specifications (MIL-F-24385F).189 Some military airfields in Europe would be in countries participating in NATO and therefore it is reasonable to assume that fluorinated AFFFs conforming to the US military specification have been used. Countries have differed in their approaches to addressing problems related to the use of AFFFs. Swedish authorities developed guidance on how to avoid PFAS containing foams and how to handle residues from firefighting foams. The state-owned airport operator Swedavia went fluorine free as of 2008.190 The Swedish Armed Forces still uses AFFFs, but only in one location (Halmstads Garnison) in which the foam and firewater is collected to avoid emissions to the environment.191 The aim is to go completely PFAS free as soon as equally good alternatives are available. The Danish Armed Forces and the Royal Danish Air Force no longer uses fluorinated AFFFs, nor do the international airports in Copenhagen and Billund.192 Only limited information has been found regarding 182 Kotterman MJJ and Kwadijk C.J.A.F (2009). PFOS onderzoek in waterbodem en vis. Report No: C064/09 (in Dutch). 183 Sportvisserij Netherlands (News) Rijnland hervat waterafvoer Schiphol terrain, 21 July 2008, Accessed: August 2018. (In Dutch). 184 Boele B, Schiphol Logistics Park blijkt vervuild, Haarlems Dagblad, 03.01.2018. Accessed 02.09.2018. 185 Schiphol Airport (2017). Annual Report (In Dutch). 186 Eurostat Transport Statistics [AIRP_TYP] Accessed 14/06/2018. 187 These numbers are approximate and do not include all EU-28 MS. Some air fields might be inactive or no-longer in existence. 188 Field J et al. (Report for Environmental Security Technology Certification Program-ESTCP) (2017). FAQs Regarding PFASs Associated with AFFF Use at U.S. Military Sites. 189 Anderson H et al. (2016). Occurrence of select perfluoroalkyl substances at U.S. Air Force aqueous film-forming foam release sites other than fire-training areas: Field-validation of critical fate and transport properties, Chemosphere 150-678e685. 190 Swedavia Airports (2018). PFAS p Swedavias flygplatser. 191 Minister of Defence Hultqvist P. (2017) Answer to question on removal of PFAS. Dnr F2017/01275/MFI 192 IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). The cost of inaction linked to PFAS exposure 71 whether fluorinated AFFFs are used by municipal rescue services or what types of AFFFs are used at other Danish airports. In Finland, the military no longer uses AFFFs during training and fluorinated AFFFs are prohibited at the fire training areas of airports.193 In Norway, the state-owned Avinor (which operates the majority of Norway's civilian airports) abolished PFOS-containing fire foam in 2001, and went fluorine free in 2011.194 The state-owned oil company Equinor has also shifted to fluorine free foams.195 An important usage area remains offshore installations. Offshore platforms accounted for 54 out of 57.6 tonnes of PFOS released in Norway up until 2005 (the amounts released from airports and fire drilling areas on land could not be estimated).196 The Norwegian offshore sector has now consciously reduced the amount of PFOS-containing foam - from 4 tonnes in 2014 to 1.1 tonne in 2017.197 In the UK, Heathrow Airport transitioned to fluorine free foam in 2012.198 Though Germany has no national restrictions on the use of AFFFs, the Umweltbundesamt (UBA) has published guidelines on the issue and some federal states have set threshold values for PFAS in foams. The number of sites in the EEA where fluorinated AFFFs may have been used are presented in Table 15. Table 15: The number of sites in the EEA where flourinated AFFFs may have been used. Sector Activity Total Aviation Main passenger airports 318 Medium passenger airports 137 Small airports no data Military airbases 239 Other fire control Fire stations 84,099 Site emergency services no data The number of fire stations across Europe is at least 85,000. However, it is not known whether all of these stations have been engaged in training involving AFFFs. The use of AFFFs, i.e., whether it is used during the trainings, accidents or both, might vary from country to country. A survey of AFFFs in Sweden by the Swedish Chemical Agency mentions firefighting training centres in Sweden, which used various types of AFFFs and notes that some types were not used after 2011 because they contained PFOS and PFOA.199 An inventory carried out in 2017 aimed to locate all non-airport fire-fighting training sites in Norway active within the past 40 years, and identified 249 previously unknown sites.200 Another possible source of contamination is release of AFFFs during efforts to extinguish fires. Figures are available for different types of fires across the EU on a yearly 193 Sahlin S (2017). PFAS in the Baltic Sea Region. 194 Avinor (2018). PFOS I Focus. (In Norwegian). 195 IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 196 Climate and Pollution Agency (2012). Inventory of PFOS and PFOS-related substances in fire-fighting foams in Norway. 197 A. Heggelund, Norwegian EPA (2018) Personal communication. 198 Ibid. 199 Swedish Chemicals Agency (2015). Survey of Fire Fighting Foam. Report 5/15. 200 Heggelund, A. Norwegian EPA. Personal communication, Sep 2018. 72 The cost of inaction linked to PFAS exposure basis (latest data from 2015) but AFFFs are specifically used to extinguish class B fires (fires involving gasoline, petroleum greases, tars, oils, oil-based paints, solvents, alcohols) and the number of class B fires could not be separated out from the total number of fires. Fluorinated AFFFs may also be present in fire suppression systems in place in industrial facilities where flammable and explosive substances are stored or used. Some studies have tried to estimate the total amounts of AFFFs found in the market (sold, or in stocks of manufacturers as well as potential users like airports, firefighting training centres or petroleum refineries). For instance, a report from 2009 estimated the quantities of PFOS containing AFFFs sold in the Dutch market in 2002 as 212,000 l and concluded that around 25 million l of foam have been placed in the Dutch market in 20 years. Some 75% of these stocks were estimated as still unused. Around 18% of the stocks were thought to have been sold to the aviation industry, and another 11% sold to the fire prevention/protection industry.201 Another study from 2008 estimated that in order to manufacture AFFFs between 1970 and 2002, 10,000 tonnes of POSF were produced/used. The amounts released into air/water were 9150 tonnes of POSF and between 91 to 460 tonnes of PFOS.202 A more recent study estimated that between 1.13 to 3.81 tonnes of firefighting foam (monomers) is used annually in the EU.203 A recent US study identified 40 novel groups of anionic, zwitterionic and cationic PFAS that had been never observed before, using non-target screening of groundwater from the areas near 13 firefighting training sites. Water samples were collected between 2011 and 2015. Their chemical composition suggested that they were either produced via electrochemical fluorination (a process of manufacturing of AFFFs by 3M which was phased out in 2002) or were compounds found in AFFFs prior to 1988, for which the composition is not known.204 It is therefore reasonable to assume that both previous and new generation AFFFs will continue to be a major source of pollution into the future.205 4.4 Case Study 4: Exposures during the use phase of PFAStreated products In order to have an overall view of the long-term costs of inaction, it is important to keep in mind the impact pathways of exposure from those products during the use phase of the PFAS life cycle. A 2000 3M study estimated that 85% of the indirect emissions of POSFs result from losses during the use and disposal stages.206 201RIVM (2009). Estimation of emissions and exposures to PFOS used in industry. Report 601780002/2009. 202Alexandre G. Paul et al, (2008). A First Global Production, Emission, And Environmental Inventory For Perfluorooctane Sulfonate, Environmental Science & Technology 2009 43 (2), 386-392. 203Ministry of Infrastructure and Environment of Netherlands and Public Waste Agency of Flanders (2016). Inventory of awareness, approaches and policy: Insight in emerging contaminants in Europe. 204 Krista A. Barzen-Hanson et al. (2017). Discovery of 40 classes of per- and polyfluoroalkyl substances in historical aqueous film-forming foams (AFFFs) and AFFF-impacted groundwater, Environmental Science and Technology. 205 Banzhaf S et al. (2016). A review of contamination of surface-, ground-, and drinking water in Sweden by perfluoroalkyl and polyfluoroalkyl substances (PFASs), The Royal Swedish Academy of Sciences. 206 Alexandre G. Paul et al. (2008). A First Global Production, Emission, And Environmental Inventory For Perfluorooctane Sulfonate. Environmental Science & Technology 2009 43 (2), 386-392. The cost of inaction linked to PFAS exposure 73 A 2015 study by KEMI provides a useful compendium of PFAS in products207 and lists a number of areas of application (see Table 16): Table 16: Products and areas of applications for PFAS Textiles and leather, including impregnating agents Paper- and food-packaging Fire-fighting foam Cosmetic products Household products Paint, printing ink and lacquer Cleaning agents and polish Non-stick products Ski wax Hard- and decorative chrome plating Hydraulic systems in the aviation industry Photographic and electronic equipment and components Photographic surface layers Photoresistors and anti-reflective coatings for semiconductors Synthesis chemicals (intermediates) Medical devices Building materials Oil and mining production Plant protection agents For many of the applications above, the primary pathways of exposure will occur during the manufacturing of the product or at end-of-product disposal. For example, the application of PFAS as an anti-vapour suppressant during chrome plating will result in releases during the manufacture of the chrome-plated product, and virtually no PFAS will be released when the final product is being used or disposed of at end of life. The use of PFAS in manufacture of medical devices will similarly have the greatest impact during the manufacturing and disposal of the product. These impact pathways are addressed in Case Study 2 on manufacturing and Case Study 4 on end-of-life disposal. However, for other products, the use phase will also have significant impacts on people and the environment, e.g., the use of PFAS in cosmetic products will have environmental impacts when the product is washed off by the user and the PFAS enters sewers, sewage treatment plants and eventually waterways. The laundering and re-impregnation of textiles and cleaning of leather treated with PFAS and the use of PFAS surfactants in cleaning agents or as agents in pharmaceuticals will likewise result in releases of PFAS to sewers and waterways. This case study focuses on three categories of products containing PFAS where the use of the product leads to human exposure and to releases to the environment, where they then accumulate. 207Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. Report 7/15. 74 The cost of inaction linked to PFAS exposure  PFAS-treated carpets PFAS-treated food contact materials cosmetic products containing PFAS. It concludes with a look at how PFAS from both domestic and industrial sources will be collected into sewers and then channelled to urban waste water treatment plants where they end up in the wastewater discharges and sewage sludge. 4.4.1 Product 1 - PFAS-treated carpets PFAS have been used to give stain, soil and oil resistance to carpets and rugs since the 1980s.208 PFAS can be applied to carpets at four different points: (1) during the manufacture of fibers for carpets, (2) during the process of manufacturing carpets and rugs, (3) at a separate finishing facility after the carpet or rug has been manufactured, or (4) post-manufacture and sale through treatment by the consumer or professional cleaner.209 Figure 4 show impact pathways for PFAS in carpets. Figure 4: Impact pathways for PFAS in carpets Source: From CA DTSC, 2018. PFAS used to treat carpets and rugs is a major source of human exposure. Surface scuffing during normal use can release PFAS on fibers in the form of tiny particulates which can be resuspended and become part of indoor air and settle into dust. Human exposure takes place when these fine particulates are inhaled or when household or office 208 California Department of Toxic Substances Control (CADTSC) (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs (Discussion draft). 209 Note that stain-resistant sprays are also sold to consumers for treating upholstered furniture and other textiles. The cost of inaction linked to PFAS exposure 75 dust is ingested. Exposure from treated carpets via ingestion is significantly higher in toddlers than adults because of greater hand-to-mouth behaviour.210 Under one scenario, this pathway was estimated to contribute 40-60% of total uptake of PFAS in infants (0-1 years), toddlers (1-4 years), and children (5-11 years). Other vulnerable populations include industrial workers, carpet installers, carpet cleaners and workers in stores selling furnishings, carpets and outdoor clothing.211 Most commercial and residential carpets sold in the U.S. are treated with PFAS, especially carpets made of synthetic materials like nylon, polypropylene, acrylic and polyester, which are prone to absorbing liquids.212 A 2000 study commissioned by 3M (a producer of PFAS treatments for carpets) estimated loss of PFAS from carpets at 50 percent over a typical lifespan of nine years, due to walking and vacuuming of the carpet.213 After 2000, improvements made by PFAS-producers and carpet manufacturers considerably reduced the loss of fluorinated treatments. After the US, the European Union is the second-biggest market in the world for carpets (both treated and untreated).214 Belgium, the Netherlands and the United Kingdom are the leading countries where carpet manufacturing takes place. Some 1.6 million tonnes of carpet are disposed of in the EU each year. Most of this ends up being incinerated or deposited in landfills. Less than 3% of carpets marketed in the EU are subsequently recycled. Side-chained fluorinated polymers are now the most commonly used carpet and rug PFAS treatment in the U.S. and in Europe. However, longer-chain PFAS continue to be manufactured in China, Russia and India, and carpets imported from those countries as well as carpets made from recycled materials may still contain long-chain PFAS. A Danish survey of rugs marketed for children found PFAS in five of the 21 carpets screened for total-fluorine in the textile surface.215 4.4.2 Product 2 - Food contact materials treated with PFAS PFAS are used in the paper industry for producing papers that resist grease and water. These are used for manufacturing food packaging, e.g. plates, popcorn bags, and 210Trudel D et al. (2008). Estimating Consumer Exposure to PFOS and PFOA. Risk Analysis: An International Journal 28.2: 251-269. 211 California Department of Toxic Substances Control (CADTSC) (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs (Discussion draft). 212 California Department of Toxic Substances Control (CADTSC) (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs (Discussion draft). 213 Battelle Memorial Institute 2000, as cited in CA DTSC, 2018. 214 Onyshko, J and Hewlett R (Anthesis Consulting Group) (2018). Toxics in carpet in the European Union. 215 Danish Environmental Protection Agency (2016). Survey and risk assessment of chemical substances in rugs for children. 76 The cost of inaction linked to PFAS exposure other packaging materials, e.g., cartons, masking tape.216 The end product may contain 1-1.5% PFAS by weight.217 PFAS in food packaging can leach into food, increasing dietary exposure.218 Information on PFAS in paper- and food-packaging is hard to come by as it can be regarded as confidential business information. However, a recent U.S. study of food contact materials (FCM) used in fast food restaurants found detectable fluorine in 46% of food contact papers and 20% of paperboard samples analysed of the 400 samples analysed for total organofluorine content.219 The presence of fluorinated chemicals in fast food packaging was seen as indicative of FCMs being a significant source of dietary PFAS exposure and environmental contamination.220 Similar levels of PFAS were found in fast food packaging gathered in Europe in 2017 and tested by the Danish Consumer Council's Think Chemicals programme.221 The use of PFAS in compostable food packaging might present an additional source of contamination for soil and vegetation.222 Contamination of composts from PFAS and related substances has been documented.223 4.4.3 Product 3 - Cosmetic products containing PFAS PFAS are used in various cosmetic and hygiene products such as sun screens, body lotions, shaving creams, dental floss and a variety of make-up products (e. g. lipsticks, eyeshadows).224, 225 The use of PFAS in cosmetics is related to their surfactant qualities which help with the penetration of the products through the skin (such as creams) or their capacity to make the skin brighter and absorb226 more oxygen. They are also used to make the cosmetic products oil- and water-repellent, and weather resistant.227, 228 Compared to other uses of PFAS such as in fire-fighting foams, research in the area of PFAS in cosmetics remains limited.229 Information on the quantities of PFAS used in 216 Trier X et al. (2011). Polyfluorinated surfactants (PFS) in paper and board coatings for food packaging. Environ Sci Pollut Res Int 18(7):1108-1120. DOI: https://doi.org/10.1007/s11356-010-0439-3 217 UNEP/POPS/POPRC.9/INF/11 2013. 218 Begley TH et al. (2008). Migration of fluorochemical-paper additives from food-contact paper into foods and food simulants. Food Additive and Contaminants: Part A 25(3):384-390. 219 Laurel A et al. (2017). Fluorinated Compounds in U.S. Fast Food Packaging. Environmental Science & Technology Letters. 220 Tittlemier SA et al. (2007). Dietary exposure of Canadians to perfluorinated carboxylates and perfluorooctane sulfonate via consumption of meat, fish, fast foods, and food items prepared in their packaging. Journal of Agricultural and Food Chemistry. 55(8):3203-3210. 221 Chemical Watch (10 March 2017)."High levels of fluorinated substances found in EU fast food packaging." 222 Center for Environmental Health (2018). Avoiding Hidden Hazards. 223 Fuchs Jacques G (FiBL) (2008). Compost and digestate: sustainability, benefits, impacts for the environment and for plant production. 224 Fujii et al. (2013). Occurrence of perfluorinated carboxylic acids (PFCAs) in personal care products and compounding agents. Chemosphere 93 (2013) 538-544. 225 Danish EPA. Risk Assessment of fluorinated substances in cosmetic products (Not yet published). 226 Ibid. 227 Fujii et al. (2013). Occurrence of perfluorinated carboxylic acids (PFCAs) in personal care products and compounding agents. Chemosphere 93 (2013) 538-544. 228 Danish EPA. Risk Assessment of fluorinated substances in cosmetic products (Not yet published). 229 Fischer S et al. (2016). Poly- and perfluoroalkyl substances on the market and in the Swedish environment. Norman Bulletin Issue 5. The cost of inaction linked to PFAS exposure 77 cosmetic manufacturing is not available and cosmetic companies usually do not disclose information on the fluorine content.230 The current body of knowledge about PFAS in cosmetics mostly stems from individual studies that surveyed a sample of products in the market to identify different types of PFAS in their composition. Exposure to toxic substances in cosmetics can occur directly from the skin. For instance, a number of studies carried out on mice and human skin suggest that PFOA penetrates both human and mice skin.231 Absorption of these chemicals through skin may not be a significant route of exposure, but absorption can increase when used on or around the eyes, posing a greater hazard.232 Although the research in this area remains very limited, it has been suggested that significant variations in absorption may occur depending on the type of PFAS used in the products, and the other chemicals present.233 One of the PFAS identified in cosmetics is PTFE, known more widely under the brand name Teflon. The Skin Deep database compiled and maintained by the U.S.-based Environmental Working Group provides ingredient lists and safety ratings for almost 75,000 cosmetics and personal care products.234 Teflon was found in 66 different products from 15 different brands. In all, 13 different PFAS chemicals were found in nearly 200 products from 28 brands, including shampoo and shaving cream. CosIng - the European Commission's public database on substances that may be found in cosmetic products - lists 76 PFAS.235 Their technical functions lend them to a broad range of uses, such as anticaking agents, emulsifiers, anti-statics, emulsion stabilizers, surfactants, film formers, viscosity regulators and solvents. Many of the products are for use on the skin and hair. Chemicals used in personal cosmetics include perfluoropolyethers (PFPE), PFOA and perfluorodecaline.236 A survey of a wide range of consumer products by the Nordic Council of Ministers published in 2017 found concentrations of TOF (total organic fluorine) over the limits of detection for many of the products. The highest concentrations were detected in two dental floss samples(the only personal care productsincluded in the sample).They both showed a310,000,000 g/m2 for TOF and 3 g/m2 and 19.5 g/m2 concentrations for PFAS respectively.237 Table 17 shows the concentrations of different PFCAs detected. 230 Ibid. 231 Franko et al. (2012). Dermal Penetration potential of Perfluorootanoic Acid (PFOA) in Human and Mouse Skin. Journal of Toxicology and Environmental Health, Part A. 75:50-62. 232 EWG Cosmetics Database, Is Teflon in your Cosmetics? Published March 14, 2018. Accessed 08.08.2018. 233 Ibid. 234 EWG Cosmetics Database, Is Teflon in your Cosmetics? Published March 14, 2018. Accessed 08.08.2018. 235Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. Report 7/15. 236 Jamberg U and Holmstrom, K (2007). Perfluoroalkylated acids and related compounds (PFAS) in the Swedish environment. 237Nordic Council of Ministers (2017). Analysis of PFASs and TOF in products. 78 The cost of inaction linked to PFAS exposure Table 17: PFCA concentrations (g/kg) detected in dental floss samples PFBA PFHxA PFHpA PFOA PFNA PFDA PFUnD A LOD238 Dental floss 1 Dental floss 2 LOD 0.05 < LOD < LOD 0.05 0.10 < LOD < LOD 0.10 0.06 < LOD 3.47 0.06 0.06 0.104 13.1 0.06 0.08 < LOD < LOD 0.08 0.08 < LOD < LOD 0.08 0.08 < LOD < LOD 0.08 PFDoD A 0.05 < LOD < LOD 0.05 PFTriA 0.05 < LOD < LOD 0.05 PFTeD A 0.09 < LOD < LOD 0.09 Source: Nordic Council of Ministers (2017). Analysis of PFASs and TOF in products. A study carried out for a master's thesis at Lund University in 2017 investigated cosmetic products to identify PFAS. Products were grouped into larger categories as sunscreen, moisturizing creams, and foundation and eye make-up. PFAS were detected in 59 products among the 1354 surveyed (4.4% of total). The brands that had products for which tests showed PFAS content included L'Oral (4 out of 41 products tested), IsaDora (8 out of 94), The Body Shop (18 out of 98), Maybelline (6 out of 72), Biotherm (1 out of 45), Lumene (22 out of 43). Large variations of PFAS concentrations among different brands were found.239 The Danish Environment Protection Agency carried out an ingredients survey based on the information found in the database of Kemiluppen app. The Kemiluppen app enables consumers to scan the barcodes of cosmetic products in Denmark, which are then assessed by the Danish Consumer Council. The survey evaluated 11,108 products that had been scanned (some products may have been scanned multiple times) and found 78 (in 20 different types of products) with fluoroalkyl substance or other fluorinated compounds listed as contents. It was also possible to gauge a product's market size, based on how many times consumers had scanned a particular product. PTFE (polytetrafluoroethylene) was present in 13 types of products that were scanned 16,641 times. C9-15 fluoroalcohol phosphate was present in four types of products that were scanned 7,826 times. Creams and lotions contained the highest number of fluoroalkyl substances (six), followed by BB/CC creams and foundations (three in each) Another study conducted by the US Breast Cancer Fund found PFOS concentrations above detection levels in six out of 17 products tested, mainly anti-aging creams, moisturizers and skin powders, all belonging to the biggest manufacturers such as L'Oral and Proctor & Gamble.240 Perfluorinated compounds have also been found in hair and skin conditioners.241 238 LOD stands for level of detection. 239 Henricsson C (Master Thesis) (2017). The Presence of PFAS in Cosmetic Products (Frekomst av PFAS i kosmetiska produkter). University of Lund. 240 Breast Cancer Fund (2015). Anti-aging secrets exposed: chemical linked to breast cancer found in skin care. 241 Nordic Ecolabelling (2018). About Nordic SAWN Ecolabelled Cosmetic Products. Version 3.3. The cost of inaction linked to PFAS exposure 79 A campaign launched by a Swedish NGO in 2017 has resulted in six global cosmetic companies pledging to phase out the use of PFAS in their products.242 The use of personal care products leads to the products, after bathing, entering into sewerage and wastewater treatment plants. This type of contamination has received little attention, but in the case of products such as shampoos and shaving creams, it might be highly relevant. For example, one study found PFCA in 89% of the sunscreens surveyed.243 This suggests that these products are potential contaminants for the aquatic environment. The research carried out for this study was not able to find information on the quantities of PFAS used in manufacturing cosmetics. It is therefore not possible to provide an overall estimate of the scope of the problem. Though data is available on the concentrations of certain PFAS in the groundwater and drinking water systems, how much of this contamination stems from personal care products is unknown. 4.4.4 Discharges from waste water treatment plants Conventional wastewater treatment is not effective in removing PFAS from waste streams.244 Thus municipal wastewater treatment plants (WWTP) are major point sources for PFAS contamination of the aquatic environment.245 The release of PFAS to the biosphere through WWTPs can be a result of both industrial activities and their domestic usage.246 Given the extreme persistence of PFAS, these substances may end up in sewage sludge applied to agricultural land and subsequently taken up into produce for human consumption. One study found that "Often PFAS concentrations increase in wastewater treatment plants as a result of biodegradation of precursors during the activated sludge process. PFOA is generally fully discharged into receiving rivers, while about half of PFOS is retained in the sewage sludge."247 Other studies have confirmed this. A 2016 study for the Swedish Environmental Protection Agency concluded that the main transport route of PFAS from societal use of goods and chemical products into the environment was via sewage treatment plants and waste management facilities, which could constitute locally significant point sources. PFAScontaining sewage from industry could also enter the environment via sewage treatment and waste management facilities. For Sweden, emissions of PFAS into the environment 242 H&M, L'Oreal, Lumene, Body Shop, Isadora and Kicks. See: Chemical Watch: Cosmetics giant L'Oral to eliminate PFASs in products, July 2018. 243 Fujii et al. (2013). Occurrence of perfluorinated carboxylic acids (PFCAs) in personal care products and compounding agents. Chemosphere 93 (2013) 538-544. 244 Adler, A and Van der Voet J (2015). Occurrence and point source characterization of perfluoroalkyl acids in sewage sludge, Chemosphere 129: 62-73. 245 Ahrens L et al. (2016). Screening of PFASs in groundwater and surface water, Uppsala Report No: 2016:2 as cited in in CA DTSC, 2018. 246 Eriksson U et al. (Report for Swedish Environmental Protection Agency) (2015). Screening of PFASs in sludge and water from waste water treatment plants. 247 Loos et al. (2009; 2010), as cited in the Concawe report (2016). Environmental fate and effects of polyand perfluoroalkyl substances (PFAS). 80 The cost of inaction linked to PFAS exposure from sewage treatment plants (336) were estimated at 70 kg/yr via water discharges and 5 kg/yr via sludge since 2004. The level in outgoing wastewater had increased since 2009, most likely due to an increase in the use of PFAS in consumer products.248 Another study from Switzerland detected PFAAs in all the sewage sludge samples collected from 20 different WWTPs in 2008 and tested a number of historical samples from 1993-2002. Concentrations of PFOS were between 15 to 600 g/kg.249 In the same vein, a study sampling sewage sludge from 45 different WWTPs in Switzerland found elevated PFOS concentrations (median 2290 g/kg). In total, these 45 WWTPs constitute approximately of the total production of sludge in the country, with 55 000 tonnes. Total quantities of PFOS at the selected WWTPs was estimated as 7.5 kg for the year 2011 (median value), which is extrapolated as 30 kg per year for the whole country. Based on these findings, the study estimated per capita emissions for Switzerland (g/pers/day) in sewage sludge, for PFOA and PFOS, as 0.2 and 4.8 respectively (median values). It is important to note that these WWTPs were selected for being close to potential industrial pollution sources and therefore the mean emissions might be overestimated.250 A study of several small rivers in Germany concluded that discharges of waste waters were the largest contributor of PFOS to surface waters. The study sampled waste water at different stages from treatment plants between 2005 and 2006.251 The WWTPs had different industries in their proximity with different daily flows. The WWTPs received inflows from domestic, industrial (breweries, tabacco, food and plastics) and commercial sources, with domestic inflows having much lower levels of PFAS (PFOS and PFOA). The study also found that PFOA is able to pass fully from the WWTP without diminishing in concentration and to find its way to rivers, while half of the PFOS is retained in the sludge. In general the study states that PFOA concentrations found are similar to those reported for US, but PFOS concentrations were higher.252 In an Austrian253 study, investigating concentrations of PFAS in effluent samples from 21 municipal WWTPs, findings indicate varying degrees of concentrations ranging from 0 to 280 ng/l for PFHxA, 10 to 220 ng/l for PFOA, and 4 to 340 ng/l for PFOS.254 A joint 2013 European study, analysed effluents from 90 European WWTPs, and concluded that despite the phasing out of PFOS, detection levels of this substance (among others) indicates an on-going release of these substances from PFAS containing products, and that release can be not solely classified as historical. On the basis of these findings, it seems reasonable to conclude that PFAS concentrations are found in the discharges of most of Europe's wastewater treatment plants. 248 Swedish Environmental Protection Agency (2016). Hgfluorerade mnen (PFAS) och bekmpningsmedel en sammantagen bild av frekomsten i miljn (In Swedish). 249Adler A and Van der Voet J (2015). Occurrence and point source characterization of perfluoroalkyl acids in sewage sludge. Chemosphere 129: 62-73. 250 Ibid. 251 Becker M et al. (2008). Perfluoroactane surfactants in waste waters, the major source of river pollution. Chemosphere 72(1): 115-21. 252 Ibid. 253 The location of sampling sites is not mentionned in the study, but we assume it is Austria since the scientists publishing the study are based at the University of Vienna. 254 Clara M et al. (2008). Emissions of perfluorinated alkylated substances (PFAS) from point sources-identification of relevant branches. Water Science and Technololgy 58(1):59-66. The cost of inaction linked to PFAS exposure 81 4.5 Case Study 5: Impacts at end of life of PFAS-treated products 4.5.1 Background and context At the end of their useful life, consumer products containing PFAS are discarded or, in some instances, recycled. If the product is a solution such as a cleaning fluid or coating, it will frequently be disposed of down a household drain where the PFAS will flow with the other fluids into a sewer system and to an urban waste water treatment plant. If the product is an article, it will end up in a municipal solid waste stream. At that point it may be diverted for recycling or material reuse, or disposed of in a landfill or via an incinerator. Globally, many municipalities and industries still rely on landfilling for final disposal of PFAS-containing products. The latest figures from the US EPA state that in 2015 53% of municipal waste was landfilled and 13% incinerated (the rest was recycled or composted).255 In the Nordic countries, landfilling household waste is done to a very low extent; e.g. 0.5% in Sweden (2017)256, 1% in Denmark (2016)257 and 3% in Norway (2017).258 Incineration is the more common method for final treatment, with Sweden incinerating around 50% of its household waste in 2017, Denmark 51% (2016) and Norway 57% (2017). The EU-28 countries as a whole are also moving from landfilling to incineration. In 1995 64% of municipal solid waste was landfilled and 14% incinerated. In 2016, 24% of household waste went to landfills while 27% was incinerated.259 Knowledge about the necessary conditions for destruction of PFAS and what happens if those conditions are not achieved is still limited. PFAS can be broken down, but only under conditions that are so harsh, e.g. incineration at very high temperatures, that they do not occur in the normal environment.260 A 2014 study for the USEPA found that a thermal reactor system operating at 1000 C was able to destroy fluorotelomer-based polymers without resulting in the formation of detectable levels of PFOA.261 However, such temperatures may not be typical. The EU rules for municipal waste incineration require a temperature of 850 0C.262 Experiments in temperatures similar to municipal waste 255 Environment Protection Agency (2018). Advancing Sustainable Materials Management: 2015 Fact Sheet. 256 Avfall Sverige (2018). Hushllsavfall - behandlad och insamlad mngd (In Swedish). 257 Danish Environmental Protection Agency (2016). Affaldsstatistikken 2016 (In Danish). 258 Statistisk sentralbyr (2018). 426 kilo avfall per innbyggjar (In Norwegian). 259 Calculated from Eurostat: Municipal waste landfilled, incinerated, recycled and composted in the EU-28, 1995 to 2016. 260 Wang Z et al. (2017). A never-ending story of per- and polyfluoroalkyl substances (PFASs)?. Environmental Science & Technology, Mar 7;51(5). 261 Taylor P S et al. (2014). Investigation of waste incineration of fluor4otelomer-based polymers as a potential source of PFOA in the environment. Chemosphere 110: 17-22. 262 Directive 2010/75/EU of 24 November 2010 on industrial emissions (integrated pollution prevention and control) (Recast), OJ L 334/17 (17.12.2010). Article 50 requires operating conditions for a waste incineration plant to reach at least 850 C for at least two seconds. 82 The cost of inaction linked to PFAS exposure incineration plants (800 0C -1000 0C) have shown that certain fluorine products remain after combustion.263 The knowledge gap as regards the fate of these substances in incineration processes is significant and an area where the knowledge base needs to improve. If a product is disposed of through landfill, the matrix (material of the product) may break down, but the PFAS will remain. Over time (and depending on the design of the landfill and the durability of the barriers between the waste and the underlying earth), the PFAS will migrate into any liquid in the landfill and drain into leachate collection systems or directly into soil and groundwater. PFAS have been detected in landfill leachate around the world.264 A Swedish study of 26 PFAS in samples from groundwater, surface water, WWTP effluent, and landfill leachate found that landfill leachates had the highest average total PFAS concentrations (487 ng/l).265 Shorter-chain PFAAs tend to be the most abundant PFAS in landfill leachate.266 Another factor raised by the study is that only a small fraction of the total number of PFAS compounds are being analysed. A screening of the total organic fluor (TOF) content in sewage sludge showed that the individually analysed PFAS compounds only accounted for a few percent of the TOF. The authors conclude that the picture is likely to be the same for landfill leachate. PFAS release from solid waste is slow, compared to the amount of PFAS manufactured and used in consumer products each year. A study of PFAS releases from carpet and clothing using a model landfill reactor found that for the most part, the releases did not take place until >200 days of operation.267 In an actual landfill, the process of breaking down the underlying substrate to release the PFAS will take much longer. Thus, the PFAS in solid waste sitting in landfills today will continue to be released into leachate for many years into the future.268 Moreover, a review of the fate and transformation of PFAS in landfills noted that ambient air around landfill sites had found elevated concentrations of PFAS compared to upwind sites used as controls, which suggested that landfills are also potential sources of PFAS in the atmosphere.269 The stability of PFAS compounds also means that they tend to remain in other materials where they have been used. PFAS used to ensure low surface tensions during plastics moulding270 are likely to remain in the plastic waste stream during materials recycling. This has implications for the circular economy. 263 Huber S et al. (Norwegian Institute for Air Research) (2009). Emissions from incineration of fluoropolymer materials. Report number: OR112/2009. 264 Fuertes I et al. (2017), Perfluorinated alkyl substances (PFASs) in northern Spain municipal solid waste landfill leachates, Chemosphere 168: 399-407. See also Hamid H et al. (2018). Review of the fate and transformation of per- and polyfluoroalkyl substances (PFASs) in landfills. Environmental Pollution, 235: 74-84. 265Ahrens L et al. (2016). Screening of PFASs in groundwater and surface water, Uppsala Report No: 2016:2. 266 Hamid et al. (2018). 267 Lang J R et al. (2016). Release of Per- and Polyfluoroalkyl Substances (PFASs) from Carpet and Clothing in Model Anaerobic Landfill Reactors. Environmental Science & Technology, 50, 5024-5032. 268 Lang J R et al. (2017). National Estimate of Per- and Polyfluoroalkyl Substance (PFAS) Release to U.S. Municipal Landfill Leachate. Environmental Science and Technology, 21;51 (4):2197-2205. 269 Hamid H (2018). Review of the fate and transformation of per- and polyfluoroalkyl substances (PFASs) in landfills. Environmental Pollution, 235: 74-84. 270 Kennebunk, Kennebunkort and Wells Water District (2017). Per and Polyfluoroalkyl Substances (PFASs) Usage (Draft for informational purposes). See also Hamid H (2018). Review of the fate and transformation of per- and polyfluoroalkyl substances (PFASs) in landfills. Environmental Pollution, 235: 74-84. The cost of inaction linked to PFAS exposure 83 PFAS are frequently applied to paperboard used for food contact products, to make waterproof and greaseproof containers. For communities with active programs aimed at diverting biodegradable materials into composting systems, it has been an unwelcome surprise to learn that compost developed from feedstocks of mixed food and yard waste and including compostable food service ware may have high levels of PFAS, with the most prevalent being the short chain PFAS (C4 - C6).271 A Swedish study found that only a small fraction of the total number of PFAS compounds are being analysed. A screening of the total organic fluor (TOF) content in sewage sludge showed that the individually analysed PFAS compounds only accounted for a small percentage of the TOF.272 The application of compost or other soil enhancers such as biosolids (sewage sludge) in which PFAS are present to agricultural soil is leading to concern that the PFAS may be taken up by edible plants and end up bioaccumulating in the food chain. A study of greenhouse lettuce and tomatoes grown in a soil amended with biosolids affected by industrial chemicals found that the plants had taken up PFAS273, and that perfluorobutanoic acid (PFBA) and perfluoropentanoic acid (PFPeA) seemed to have bioaccumulated to a degree. The short-chain PFAS seemed to be less attached (more mobile) to the agricultural soil and to have higher crop uptake potential. The study indicates that plants grown on soil amended with sewage sludge containing PFAS can bioaccumulate PFAS, with the extent of the bioaccumulation varying depending on the concentration of PFAS, the properties of the soil, the type of crop, and the chemical under analysis. The finding that short-chain PFAS had the highest potential to bioaccumulate in produce was duplicated in a recent Minnesota study.274 The study investigated whether home gardens irrigated with PFAS-contaminated water would result in contaminated produce. Among the conclusions: PFAS in water can enter into the food chain under real-world conditions. Short-chain PFAS in water had a greater impact on levels in produce than long-chain PFAS in soil. Finally, PFAS concentrations varied according to plant part with florets found to have the highest concentrations of PFAS. 4.5.2 Cases of contamination Arnsberg, Germany In 2006, a high level of PFAS contamination was detected in the conjunction of the rivers Rhine, Ruhr and Moehne as well as nearby public water supplies. This had consequences 271 Lee L and Trim H (2018). Summary Sheet Evaluating Perfluoroalkyl Acids in Composts with Compostable Food Serviceware Products in their Feedstocks. 272 Avfall Sverige (2018). Hushllsavfall - behandlad och insamlad mngd (In Swedish). 273 Blaine AC et al. (2013). Uptake of Perfluoroalkyl Acids into Edible Crops via Land Applied Biosolids: Field and Greenhouse Studies. Environmental Science & Technology. 47(24), pp. 14062-14069. 274 Scher D et al. (2018). Occurrence of perfluoroalkyl substances (PFAS) in garden produce at homes with a history of PFAScontaminated drinking water. Chemosphere 196: 548-555. 84 The cost of inaction linked to PFAS exposure for the whole Ruhr valley catchment and Lake Moehne, which supply 5-6 million people.275 An estimated 40,000 people were exposed to drinking water that was contaminated from this source.276 The water was contaminated with PFOA, perfluorohexanoate (PFHxA), perfluorohexane sulfonate (PFHxS), perfluoropentanoate (PFPA) and perfluorobutane sulfonate (PFBS). 110.5 kg of PFAS were calculated to have entered Lake Moehne.277 This was gradually released into the Ruhr and Rhein, to make its way to the North Sea. The source of the contamination was PFAS contaminated sludge containing industrial waste, which was sold under the name of "bio-solids" and was applied at farmland at the head of the Moehne valley.278 Several actions were taken to manage the contamination. First, a monitoring system was put into place to assess the level and spread of contamination in drinking water. Several biomonitoring studies were launched to measure the pathways into mothers and children (including breastmilk), and men, as well as the Activated carbon filters were installed in water works. Recommendations were developed to reduce the consumption of fish.279 Carbon filters were installed in July 2006. The carbon filters were reactivated about every 6 months. The reactivation was undertaken in specialized centers that involved treatment of the filters in a furnace heated up to over 800 degrees Celsius. A biomonitoring study was undertaken to determine the level of blood concentration in a sample of men, women and children.280 In total, 138 children, mothers and men participated in the study. Measurements were taken before and in subsequent years after the installation of activated charcoal filtering system to remove PFOA from the drinking water. Repeat blood samples were analysed for about 20 to 25 individuals for each group (children, mothers and men). Notable decreases were detected over the two year period (see Table 18). For example, the serum concentration level of PFOA in children decreased 39%. The study suggest that the reduction may be due in part to a reduced consumption of fish from local sources. Table 18: Relative reduction of PFC plasma levels (%) between 2006 and 2008 PFOA PFOS PFHxS Children Mothers Men 39.2 (31.6-48.5) 39.4 (33.5-46.3) 25.5 (21.3-30.5) 20.1 (14.7-27.6) 21.7 (16.0-29.4) 25.0 (21.5-29.0) 18.7 (10.4-33.7) 29.6 (24.7-35.4) 14.3 (10.4-19.7) 275 IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 276 Hlzer J et al. (2009). One-year follow-up of perfluorinated compounds in plasma of German residents from Arnsberg formerly exposed to PFOA-contaminated drinking water, International journal of hygiene and environmental health. 212(5), pp.499-504. 277 Weber R (2016). Presentation for Science and Policy of Organohalogens pre-Dioxin Symposium. 278 Skutlarek D et al. (2006). Perfluorinated surfactants in surface and drinking waters. Environmental science and pollution research international, 13(5), p.299. IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 279 WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. 280 Brede E et al. (2010). Two-year follow-up biomonitoring pilot study of residents' and controls' PFC plasma levels after PFOA reduction in public water system in Arnsberg. International journal of hygiene and environmental health, 213(3), pp.217-223. The cost of inaction linked to PFAS exposure 85 As of 2006, 100 million Euros had been spent on investments in the regional water works.281 Charges were pressed against the German company providing the sludge, which declared bankruptcy as a consequence of the contamination scandal, and the CEO was taken to court. The following year, national monitoring activities by the competent authorities were initiated. Baden-Wuerttemberg, Germany Following the Arnsberg case (see previous case study), the state of Baden-Wuerttemberg analysed samples from 41 locations potentially subject to PFAS contamination in 2006.282 Measure points with elevated levels were followed up subsequent years. In 2013, PFAS was found during a routine analysis in a well belonging to Landkreis Rastatt's drinking water supply.283 Further investigations unravelled a contamination situation of unprecedented dimensions, making it the greatest contamination case in Germany both in terms of surface affected and the complexity of contaminant composition.284 As of August 2018, 644 hectares of soil in Landeskreis Rastatt and Stadtkreis Baden-Baden, as well as 240 hectares in Mannheim, are expected to be contaminated by PFAS.285 Below the area affected runs one of the larges underwater rivers in Europe, the Oberrhein-Aquifer, adding to the level of graveness of the pollution incident. Although the reason for contamination remains somewhat disputed, the explanation which mainly is put forward is the use of compost blended with contaminated paper mill waste which was applied on agricultural land between 2005 and 2008.286 An additional source would be a fire extinguishing event in 2010. The total amount of PFAS which entered the environment is hard to appreciate; an (uncertain) estimate received from a member of staff at Landkreis Rastatt reads 1,000-5,000 kg.287 Following the PFAS discovery, two waterworks providing Rastatt drinking water were taken out of use due to PFAS in the water, leaving Rastatt with one single waterwork available. To ensure safe drinking water, the local water company Stadtwerke Rastatt has invested millions of Euros in new infrastructure, groundwater monitoring, treatment methods such as active carbon filters or reversed osmosis, and in securing alternative sources of water.288 This has enabled one of the waterworks to reopen in Februari 2018. The reconstruction of pipes, water works, and installation of activated carbon filters have cost the company EUR 3.6 million by the end of 2017. The same activities are estimated to cost the company another EUR 6.2 million during 2018-2020 281 Weber R (2016). Presentation for Science and Policy of Organohalogens pre-Dioxin Symposium. 282 LUBW (2018). Grundwasserberwachungsprogramm Ergebnisse der Beprobung 2017 (in German). 283 Regierungsprsidium Karlsruhe (No date). berblick zur PFC-Problematik in Mittel- und Nordbaden. 284 Regierungsprsidium Karlsruhe (2017). Antworten auf hufig gestellte Fragen zur PFC-Belastung im Landkreis Rastatt und den Stadtkreisen Baden-Baden und Mannheim Stand August 2017. 285 Ministerium fr Umwelt, Klima und Energiewirtschaft Baden-Wrttemberg (2018). Neuer Erlass schreibt Beurteilungsgrundlage fr mgliche PFC-Belastungen von Grund- und Sickerwasser fort. 286 Klatt P and Frey A (2016). Woher kam das Zeug blo?, Frankfurter Allgemeine, 4 September 2016. Accessed September 2018. 287 R. Shlmann (2018). Landratsamt Rastatt, personal communication 288 Stadtwerke Rastatt (2018). So schtzen wir unser Trinkwasser in Rastatt (website in German) and personal communication 86 The cost of inaction linked to PFAS exposure (much depending on how frequently the carbon filter will need to be changed). These costs have so far been borne by the company without the contribution of any state funds, and thus the water price for the consumers in Rastatt has been raised from EUR 1.64 to EUR 1.86 per m3 in the course of 2017. As regards the further securement of uncontaminated drinking water, there has been an ongoing discussion between the main water company (Stadtwerke Rastatt) and the responsible authorities at municipal and state level concerning the next steps to take. Though the water company called for immediate action in order to prevent a worsening of the situation, the public authorities proposed to spend the year up to 2021 to improve the knowledge base, and to await further action until then.289 In the meanwhile, efforts to assess the extent and potential development of the contamination profile are being taken. To improve knowledge on health effects, the state and local health authorities are carrying out a blood sampling study of the population, the results of which are scheduled to be presented at the end of 2018.290 The study has cost around 257,000 Euros during 2017-2018, which includes costs for meetings with the experts, laboratory costs and the time spent on planning, carrying out and assessing the study. Two repetitions of the study are planned: one in 2020 and one in 2023, with an estimated budget of 408,000 Euro. In the region in question, growing vegetables and fruit is a major part of the local economy. Today, even certain farmers who did not use the contaminated compost are unable to grow and sell their goods, as the PFAS has spread to reach their land as well.291 To ensure that no produce with unacceptable levels of PFAS reach the consumers, the state of Baden-Wurttemberg has integrated a pre-harvest monitoring program targeting PFAS into their foodstuff monitoring framework. This is financed by the state and its costs have already amounted to more than one million Euros.292 The state Environmental Department (LUBW) has also initiated pilot studies of PFAS in groundwater, focusing on potential contamination deriving from agriculture. As a result, PFAS were included into the regular state water monitoring program from 2015, with a planned timeline of four years. As of 2017, around 50% of the measure points showed PFAS contamination (around half of them at levels below 10 ng/l). Monitoring in general is likely to prove necessary for decades to come. No clear verdict in terms of responsibility has yet been reached. Nevertheless, a verdict from Mannheim administrative court required the party who provided the contaminated compost to farmers in the region to pay for the preliminary investigation of the soils (35,000 Euros).293 The court considered his provisioning of the papermill waste compost 289 Klatt P and Frey A (2016). Woher kam das Zeug blo?, Frankfurter Allgemeine, 4 September 2016. Accessed September 2018. 290 Gesundheitsamt Rastatt (No date). Informationen zum Stand der PFC-Blutkontrolluntersuchung im Landkreis Rastatt. 291 Klatt P and Frey A (2016). Woher kam das Zeug blo?, Frankfurter Allgemeine, 4 September 2016. Accessed September 2018. 292Ministerium fr Umwelt, Klima und Energiewirtschaft Baden-Wrttemberg (No date). PFC-Belastung in Baden-Baden und im Landkreis Rastatt. 293 Shlmann, R. (2018), Landratsamt Rastatt. Personal communication. The cost of inaction linked to PFAS exposure 87 mix to farmers as a cheap way of disposing of industrial waste, rather than providing useful soil improver. For the rest, the state has covered most of the costs related to the initial investigations and measures. 900,000 Euros has been spent on modelling groundwater remediation, and 1.5 million Euros on other research projects.294 As regards public spending, 4.4 million Euros of state funds and almost 1.7 million Euros of municipal funds have so far been used in relation to the PFAS contamination. How much expenses are yet to come remains unknown, but it has been estimated that a complete remediation of the affected soils might amount to 1-3 billion Euro.295 A cost analysis dating from 2015 estimated that each hectare would cost 5.5 million Euros to remediate. With today's 640 contaminated hectares in Landkreis Rastatt, this would mean a total cost of around 3.5 billion Euro. To actually perform this enormous soil exchange is however not realistic, partly due to the vast quantities of soil that would be "lost".296 In addition, the 2015 study estimated that the cost of groundwater remediation would amount to 150 million Euros. As today even more PFAS are likely to have reached the groundwater, it is probable that this figure will not suffice. Rockford, Michigan The leather tannery complex owned by the footwear company Wolverine World Wide treated leather for manufacturing into shoes sold under the brand HushPuppy. PFAS purchased from the company 3M were applied to the leather (along with other chemicals) and the shoes were marketed as being waterproof. The chemicals were stored in drums at outdoor locations. A 2000 summary of hazardous chemicals on site stated that 16,590 pounds (7,525 kg) of Scotchgard FC-3573 and 64,409 pounds (29,215 kg) of Scotchgard FX-3573 were kept on site. In the early 2000s, 3M reformulated its product (Scotchgard) to remove PFOS from the formulations because of evidence of the chemical's toxicity, bioaccumulability and persistence in the environment. High concentrations of both PFOA and PFOS have now been found in soil and groundwater at the now unused manufacturing site, and the chemicals have now migrated into a creek that runs into the nearby Rogue River.297 Testing of surface foam at a local dam also found very high levels of PFAS, as did testing of fish caught in the Rogue River. The State of Michigan has now issued a fish consumption advisory to warn local anglers not to eat fish from the most contaminated stretch of the river. In 2017 local residents alerted environmental authorities to a former licensed disposal facility owned and operated by Wolverine as well as several unregulated dumpsites where leather scraps and other manufacturing waste were deposited, and requested testing of nearby wells.298 In May 2017, PFAS contamination was detected, and Wolverine began a more extensive well sampling program. It also started to provide bottled water and water filters for affected households. 294Ministerium fr Umwelt, Klima und Energiewirtschaft Baden-Wrttemberg (No date). PFC-Belastung in Baden-Baden und im Landkreis Rastatt. 295 IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous film-forming foams (AFFF). 296 Shlmann, R. (2018). Landratsamt Rastatt, personal communication. 297 See State of Michigan website on its PFAS response for more information. 298 Ibid. 88 The cost of inaction linked to PFAS exposure Later in 2017, the State of Michigan developed a groundwater criterion for PFOA and PFOS of 0.07 g/l (70 ppt) for protection of drinking water. In one of the areas close to the now closed landfill, 561 homes were tested for PFAS in drinking water; PFOA and PFOS was detected in 197 homes, 70 of which had concentrations over 70 ppt. The highest concentration found was 62,500 ppt. In another affected area, 690 homes were sampled, 391 had detected levels of PFOA and PFOS, 38 homes had concentrations over 70 ppt, and the highest concentration was 49,200 ppt.299 Blood tests of residents have shown high levels of PFAS. One long-term resident whose well tested at 38,000 ppt learned that four different PFAS were found in her blood for a combined total of 5 million ppt. One chemical, PFOS, was found at 3.2 million ppt - about 750 times the national blood level average of 4,300 ppt.300 While it is not possible to pin specific health impacts to a specific exposure of PFAS, the resident knew of epidemiology studies that had found links to the thyroid problems she suffered from and the liver cancer from which her husband had died in 2016. 4.5.3 Extent of the exposure due to PFAS disposal to land A study301 of US municipal landfills calculated that the total mass of PFAS from landfill leachate to wastewater treatment plants in 2013 was between 563 and 638 kg. The researchers measured concentrations of 70 PFAS in 95 samples of leachate from 18 landfills in the USA of varying climates and deposit ages, then linked estimates of total annual leachate volumes in the US with the concentrations measured for the 19 PFAS where >50% of samples had quantifiable concentrations. Participating landfills were publicly owned, so they contained mainly municipal solid waste and some sewage sludge, but probably did not receive industrial waste. FTCA (5:3 fluorotelomer carboxylic acid) was the dominant PFAS in the majority of samples collected. The ~600 kg/yr estimate for total PFAS mass release in US landfill leachate in 2013 has limitations. It is based on PFAS concentrations from 18 sites around the US, and then extrapolated to cover the total of 1540 landfills in the US that year. While none of the landfills in the study reported accepting waste from PFAS, textile or carpet production, some municipal solid waste landfills may accept such waste. For example, PFOS concentrations from leachate at a landfill that received wastewater treatment sludge from a 3M facility in Minnesota302 were measured at 136 micrograms/L, compared to the ranges of concentrations (2-29 micrograms/L) in the leachate from the 18 landfills sampled in the study. In addition, it does not account for historic landfills (the US had some 6000 landfills in 1988) that may be unlined and 299 Ibid. 300 Ellison G (2018). Belmont woman's blood is 750 times national PFAS average, Mlive, 11.01.2018 Accessed 15.10.2018. 301 Lang JR et al. (2017). National Estimate of Per- and Polyfluoroalkyl Substance (PFAS) Release to U.S. Municipal Landfill Leachate. Environmental Science and Technology, 21;51(4):2197-2205. 302 Oliaei F et al. (2013). PFOS and PFC releases and associated pollution from a production plant in Minnesota (USA). Environmental Science Pollution Research. 20: 1977-1992. The cost of inaction linked to PFAS exposure 89 closed without low permeability covers.303 These unlined landfills will continue to pose a risk of PFAS leakage to groundwater for many years to come. The following examples illustrate that municipal and industrial landfills are sources of legacy PFAS, primarily affecting groundwater and run-off surface waters: a 2010 study calculated around 90 kg/yr for 44 PFAS in treated leachate from all (~1700) landfills in Germany304; a 2016 study of PFAS in the Swedish environment from different sources estimated 70 kg/year of PFAS emissions via leachate from 365 landfills.305 Except for 8 kg of PFAS spread to land, the leachate was sent for sewage treatment; a 2017 study of 4 municipal waste landfills across northern Spain gathered data on the occurrence and concentration of 16 PFAS in the leachate from those sites. The landfills served 1.8 million people. Based on the volume of leachates from the landfill sites, it was estimated that the combined discharge of the 16 PFAS was 1.2 kg/year.306 The costs of remediation for removing PFAS contamination from affected waters would therefore be the same as for other sources of PFAS contamination. 4.6 Other costs related to PFAS contamination Some of the non-quantifiable costs of exposure to PFAS are the experiences of the individuals and communities affected, i.e., as they come to understand that the drinking water they have been consuming has contained a contaminant that may result in health impacts that do not become evident until years hence. Those experiences are part of the overall story concerning the socioeconomic costs of PFAS. "For years, folks trusted that their water providers were delivering a completely safe product and knew exactly what was in it. At the same time, the providers trusted that regulators were adequately protecting water sources and knew exactly what was in our rivers. That trust is gone." Editorial in Wilmington NC Star News, 24 June 2018307 303 Lang JR et al. (2017). National Estimate of Per- and Polyfluoroalkyl Substance (PFAS) Release to U.S. Municipal Landfill Leachate. Environmental Science and Technology, 21;51(4):2197-2205. 304 Busch J (2010). Polyfluoroalkyl compounds in landfill leachates. Environmental Pollution, Vol 158:5 , 1467-1386. 305 Swedish Environmental Protection Agency (2016). Hgfluorerade mnen (PFAS) och bekmpningsmedel en sammantagen bild av frekomsten i miljn. 306 Fuertes I et al. (2017). Perfluorinated alkyl substances (PFASs) in northern Spain municipal solid waste landfill leachates. Chemosphere 168: 399-407. 307 Editorial, We must take control of our water quality, Starnews Online, 25.06. 2018. Accessed 11. 09. 2018. 90 The cost of inaction linked to PFAS exposure Other more quantifiable costs of inaction may include loss of property value, reputational damages to a polluting company, (e.g, the recent Miteni bankruptcy), court awarded damages and financial settlements. For example, the Australian government announced AU$73.1 million budgeted to support those affected by PFAS contamination, of which AU$55.2 million will be spent across five years to give people access to safe drinking water.308 The government has already spent more than AU$100 million on PFAS, and has not yet started any remediation work. Some of the measures taken have addressed the human impacts of the contamination, e.g., public outreach, a help desk, and counselling services for affected communities. Table 19 summarises a number of indemnities paid in legal settlements by companies with facilities that led to PFAS contamination of the environment as well as PFAS-related health effects. Table 19: Amounts of some indemnities paid in legal settlement in relation to PFAS contamination due to different sources Incident Year1 Description Amount (EUR)2 Contamination of drinking water in Hoosick Falls NY due to industrial production (US) Contamination of natural resources in Minnesota due to the 3M plant (US) Contamination of water supplies in Ohio (US) Contamination of Hyannis MA water supply due to AFFFs (US) Contamination of agricultural fields in Baden-Wuerttemberg 2014-Ongoing 2018 2017 2017 2013 Legal case on-going between the municipality and Saint-Gobain Performance Plastic and Honeywell International (as a part of on-going fees to be paid to the municipality) Legal settlement between 3M and the State of Minnesota Agreement reached between Dupont Chemicals and Chemours in class action law suit with the residents Legal settlement between the town of Barnstable and Barnstable country Initial investigations payed for by compost salesman who had provided contaminated compost to farmers 443,000 710,400,000 595,000,000 2,617,000 35,000 Note: 1) Year might refer to year of detection, or the year costs have incurred. 2) Costs in other currencies are converted to Euro, using average annual rates for the year they incurred. 4.7 Summary of case study findings As the five case studies illustrate, PFAS are released to the environment from many sources, from production and manufacturing plants to specialist uses such as AFFFs for firefighting and everyday consumer products such as clothing, pizza boxes and cosmetics. 308 Swanson M, People, not politics, must be PFAS priority, The Herald, May 2018. Accessed 12.11.2018. The cost of inaction linked to PFAS exposure 91 The first three case studies cover the activities that account for a large proportion of the PFAS released directly into the environment. Case Study 1 looks at how the industrial facilities producing the fluorochemicals and fluoropolymers, while relatively limited in number, are significant individual emitters of PFAS into the air, soil and waterways. The study estimates that 12 to 20 facilities actively produce fluorochemicals in Europe, that these facilities are significant sources of PFAS released to the environment, and that exposure to workers at these plants is high. The study did not identify fluorochemical production facilities in the Nordic countries. Other industrial activities with the potential to release PFAS to the environment take place throughout Europe, including the Nordic region. Case Study 2 considers the manufacture and commercial use of PFAS-containing products, including textile and leather manufacturing; metal plating, including chromium plating; paper and paper product manufacturing; paints and varnishes; cleaning products; plastics, resins and rubbers; and car wash establishments. Releases of PFAS occur via the air or effluent entering sewerage and wastewater treatment plants, before discharge into waterways. The case study gathers Eurostat figures for the number of large companies and SMEs carrying out the industrial activities reviewed. In the absence of information concerning how many of the companies use PFAS in their manufacturing, a range of 3% to 10% of facilities is suggested. The third major source of direct emissions is the widespread use of aqueous film-forming foams (AFFFs) used to extinguish fires in emergencies or during training, especially around airports and military bases. Where the AFFFs have migrated to groundwater and other sources of drinking water, nearby communities have been affected by elevated levels of PFAS in their drinking water. It is noted that other uses of AFFFs for firefighting, especially at major industrial facilities, may also be a significant source, but one that has so far received little attention. High performance non-fluorinated AFFFs are now available, but legacy emissions from PFAS in AFFFs used in the past will continue to be a problem for years to come. The two case studies on the use and end-of-life phases of consumer products account for the remaining releases. They can be direct as well as indirect sources of exposure to PFAS. A 2000 study carried out for 3M estimated that 85% of the indirect emissions of POSFs (a precursor of PFOS) would result from losses during the use and disposal stages. More recent information on the proportion of indirect emissions of PFAS during the use and disposal stages of the chemicals' life cycle was not found. Case Study 4 considered PFAS-treated carpets, PFAS-treated food contact materials, and cosmetics as examples of how a product's use is likely to lead to human exposure through ingestion and dermal absorption. Consumer products can also lead to releases to the environment when the product is washed off or laundered, entering sewers, treatment plants, and eventually waterways. The availability of suitable non-fluorinated alternatives makes the use of PFAS in many of these products unnecessary. Case Study 5 looks at end-of-life impacts of PFAS-treated products. Waste incineration may destroy PFAS in products if 1000 C operating temperatures are reached, but such temperatures are not typical of most incineration capacity (the EU's Industrial Emissions Directive, for example, requires a temperature of 850 C). If landfilled, the PFAS will remain even after the product's core materials break down. The compounds will 92 The cost of inaction linked to PFAS exposure eventually migrate into liquids in the landfill, then into leachate collection systems or directly into the natural environment. They may then enter drinking water supplies, be taken up by edible plants and bioaccumulate in the food chain. A number of other costs related to PFAS contamination include loss of property value, reputational damage to a polluting company (as in the case of the recent Miteni bankruptcy), and the costs incurred by public authorities in responding to affected communities - including public outreach, surveys of contamination, and remedial measures. For future investigations of this nature, it would be useful to have more information concerning the sites where production of PFAS and/or where manufacturing of products involving PFAS is occurring - both current and legacy activities. National inventories of such sites, including where fluorinated AFFFs have been used, would help estimates of the numbers of affected populations, and the extent of contamination where remediation may be needed. Another suggestion is to include industries producing or using PFAS in the European Pollutant Release and Transfer Register, so that information on the location and amount of releases to air and to water is available. National registries of products containing PFAS would help inform how PFAS are used and contribute to better characterisation of the major sources of exposure from products. Finally, more research is needed concerning what happens to PFAS discharged from wastewater treatment plants and during incineration of PFAS. The cost of inaction linked to PFAS exposure 93 94 The cost of inaction linked to PFAS exposure 5. Estimates of costs of inaction linked to exposure to PFAS 5.1 Health-related costs of exposure to PFAS This section presents findings for the health-related costs of PFAS exposure at three different levels - occupational (high) exposure, elevated (medium) exposure and background (low) exposure. The quantification was carried out for a selection of scenarios and health endpoints. Health-related costs that could not be quantified are reviewed qualitatively. 5.1.1 Occupational (high) exposure scenario: PFAS production and manufacture of PFAS-containing products Individuals who are regularly exposed to PFAS through their occupation (e.g. workers in manufacturing plants producing PFAS or PFAS-treated products) may face greater risk of developing illnesses that affect their health and well-being. The desk research identified two studies that investigated elevated health risks due to PFAS exposure among workers in Europe. One study investigated serum concentrations of PFOA and liver enzymes from 56 workers in a fluorochemical production plant. The study found that PFOA serum concentration was associated with higher ALT, GGT and ALP enzymes and lower bilirubin.309 Another 2001 study conducted by a staff epidemiologist from 3M reported that occupational exposure to PFOA and PFOS in chemical plants in Antwerp and another site in the United States was associated with an increased level of cholesterol and triglycerides.310 However, the findings were more cautiously stated in the study published several years later.311 One of the most well-known studies of occupational exposure to PFAS was carried out under the C8 Health Project, which gathered data from workers from the Dupont Washington Works facility in West Virginia from 1952-2008. The sample of workers was known as the worker cohort while the C8 Health Project also gathered data from a community cohort. A multi-mortality study was conducted for this worker cohort, which 309 Costa G et al. (2009). Thirty years of medical surveillance in perfluorooctanoic acid production workers. Journal of Occupational and EnvironmentalMedicine 51:364-372. 310 Olsen G W (2001). A longitudinal analysis of serum perfluorooctanesulfonate (PFOS) and perfluorooctanonate (PFOA) levels in relation to lipid and hepatic clinical chemistry test results from male employee participants of the 1994/95, 1997 and 2000 fluorochemical medical surveillance program. Final report Epidemiology. 311 Olsen G W et al. (2003). Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations. Journal of Occupational and Environmental Medicine, 45(3), pp.260-270 and Lerner, S. (2018), 3M knew about the dangers of PFOA and PFOS decades ago, internal documents show, The Intercept, 31.07.2018. The cost of inaction linked to PFAS exposure 95 had a target population of 6,026.312 The study found evidence of a potential elevated risk of death from kidney cancer due to prior exposure to PFOA in the workplace. Other sources suggest that occupational exposure to PFAS may have developmental toxicity. For example, a pregnant female worker who had direct contact with PFOA through her employment at the Dupont facility in West Virginia gave birth to a child with multiple birth defects. Soon afterwards, the company did not allow women of reproductive age to work directly with the chemical.313 This endpoint, however, was not confirmed in the C8 Health Project in West Virginia.314 Another study conducted among workers from a 3M plant in Decatur, Alabama did not find a relationship with pregnancy outcomes.315 The small number of individuals in this category of pregnant workers may limit the detection of a statistical pattern between exposure to PFOA and pregnancy outcomes. Overall, the evidence suggets that occupational exposure to PFAS may lead to an elevated risk of kidney cancer, a disease that can lead to significant costs in terms of health care expenditures, reduced quality of life, and premature death. The valuation for the health-related costs due to occupational exposure to PFAS focused on the kidney cancer endpoint for which there is epidemilogical evidence. An assessment was made by drawing on the findings from the cohort mortality study of workers from West Virginia316 extrapolated to the European context. Like other epidemiological studies carried out under the C8 Health Project, the cohort mortality study had a robust design and was overseen by a Science Panel of three epidemiologists. The sample for the study was restricted to individuals who had worked at least one day at the Dupont chemical plant in West Virginia between 1948 and 2002. The sample was divided into four groups based on their estimated level of exposure to PFOA. The study found that high occupational exposure to PFOA was associated with an increased risk of death due to kidney cancer. As explained in the Chapter 4, it was not possible to develop a firm estimate of the number of plants that manufacture PFAS or products using PFAS in Nordic and EU countries. Some assumptions were made in order to generate estimates to support the valuation of the health-related costs for the study. The first assumption is that 20 manufacturing plants produce PFAS in European countries (see conclusion to Case Study 1 at page 64). The second assumption relies on Case Study 2's research that identified a total of 352,764 small (less than 250 employees) and 780 large manufacturers (more than 250 employees), which may use or emit PFAS, in EEA countries. As Case Study 2 explains, no information was available concerning which plants used PFAS in manufacturing their products nor how many workers were employed at each plant. In the absence of concrete 312 Steenland K and Woskie S (2012). Cohort mortality study of workers exposed to perfluorooctanoic acid. American journal of epidemiology. 176(10) pp.909-917. 313 Mordock J, C8 suspected in birth defects: One woman's story, Delaware Online, April 2 2016. Accessed 12.10.2018. 314 An overview of the probable link evaluations from the C8 Health Project can be found here 315Grice M et al. (2007). Self-reported medical conditions in perfluorooctanesulfonyl fluoride manufacturing workers. Journal of Occupational and Environmental Medicine 49:722-729. 316 Steenland K and Woskie S (2012). Cohort mortality study of workers exposed to perfluorooctanoic acid. American journal of epidemiology. 176(10) pp.909-917. 96 The cost of inaction linked to PFAS exposure information, the calculations assume that 3% to 10% of these plants use PFAS. As a last step, in order to generate the size of the exposed population, it was assumed that small plants had 30 workers on average while large plants had 300 workers on average. The population with occupational exposure in Europe was therefore estimated to range between 334,508 and 1,091,692 (see Annex 2 for more information). The analysis assumed that these individuals suffered an elevated risk of mortality from kidney cancer as documented from the West Virginia study.317 The number of additional deaths due to PFAS exposure in this population was then estimated to be between 3.6 and 11.8. These deaths were then monetised using the lower bound of the ECHA "value of a statistical life". Table 20 presents a summary of the findings, which give an indication as to the potential scale of the health impacts for the scenario of occupational exposure to PFAS. A key uncertainty in constructing these estimates was the number and distribution of worker exposure to PFAS. Table 20: Occupational exposure scenario: Monetised annual health impact for manufacturing worker exposure to PFAS Findings Annual estimates Exposed population in Europe Population experiencing elevated health risk Deaths due to kidney cancer linked to PFAS exposure Value of life lost 335 thousand to 1.1 million 83,627-273 thousand 3.6 to 11.8 lives lost EUR 12.7 million-41.4 million Note: For more information on this calculation please refer to Annex 2. 5.1.2 Elevated (medium) exposure scenario The case studies identify two populations that are at risk for elevated exposure to PFAS. The first population are communities that are in close proximity to chemical plants that produce PFAS or PFAS-treated products (Case Studies 1 and 2). Contaminated water from these plants may enter the drinking water system serving the communities. The second population are communities that live close to sites contaminated by aqueous fire-fighting foams (Case Study 3). PFAS in the foam can seep into the ground and groundwater, leading to contamination of local drinking water supplies. A Swedish study found that up to 300,000 of residents in the country - or about 3% of the total population - are or have been exposed to levels of PFAS above the action value, via municipal drinking water.318 The study highlights two main sources for the contamination: close proximity to a plant producing PFAS or PFAS-treated products, and close proximity to areas with high utilisation of aqueous fire-fighting 317 Steenland K and Woskie S (2012). Cohort mortality study of workers exposed to perfluorooctanoic acid. American journal of epidemiology. 176(10) pp.909-917. 318 Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). It should be noted that 1.4 million of Sweden's 9.9 million residents are not connected to a municipal water system. It is assumed nonetheless that all residents have contact with municipal waste water at least occasionally. The cost of inaction linked to PFAS exposure 97 foams. These same individuals may also suffer higher exposure to PFAS due to contaminated surface water or air emissions. The case studies provide other estimates for the number of individuals with elevated exposure. For example, Case Study 1 notes that 750,000 individuals living downstream from the Dordrecht Chemours plant had elevated exposure to PFAS via air and drinking water, and were advised to limit the consumption of vegetables grown in the area.319 On the other hand, the affected population may be minimal if the surrounding vicinity has few inhabitants. For example, in the case of the 3M manufacturing plant in Antwerp, the PFAS contamination was considered to have concentrated on the area of the port where few people reside. Elevated levels of PFAS in affected communities is highlighted in several studies (see Table 21). The blood serum concentration levels and the type of PFAS compounds, however, vary. Studies from the C8 Health Project in West Virginia found serum concentration levels of PFOA at 350 ng/ml among the nearby community.320 The serum concentration level of PFOA in the general population in the United States was estimated to be 3.07 ng/ml in 2010.321 A study from a case of aqueous fire-fighting foam contamination in Sweden also found an elevated level of PFAS exposure among residents nearby.322 Blood samples were taken from 3,660 persons, of whom 3,412 were in the contaminated area (Ronneby) and 242 people from a reference population in a nearby community (Karlshamn) that did not receive the contaminated water. Of the seven PFAS compounds identified in the contaminated water, three were identified at elevated levels in the blood serum of the sampled population and also elevated levels in the groundwater (above 90 ng/ml). The three main PFAS compounds identified in the blood serum were PFHxS, PFOA, and PFOS. 319No Author, Another Chapter: Chemours releases GenX and PFOAs into waters globally, Encore, 08.08.2018. Accessed 10.10.2018. 320 Frisbee S J et al. (2009). The C8 health project: design, methods, and participants. Environmental health perspectives, 117(12), 1873-82. DOI: https://doi.org/10.1289/ehp.0800379 321 Center for Disease Control and Prevention (2018). Fourth National Report on Human Exposure to Environmental Chemicals. Updated Tables Volume1. 322 Li Ying et al. (2017). Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. 98 The cost of inaction linked to PFAS exposure Table 21: Review of serum concentration levels (ng/ml) in contaminated and outside contaminated areas - three examples Ronneby, Sweden (AFFF contamination) West Virginia, USA (chemical plant) Veneto Region, Italy1 (chemical plant) Sample size Study year PFHxS PFOA PFOS Outside contaminated area 242 Contaminated area 3,418 2014-2016 0.84 152 1.59 10.4 4.21 176 Outside contaminated area2 Contaminated area3 2000+ 32,254 2009-2014 1.35 3.07 4.99 2005-2006 n.a. 32.9 19.6 Outside contaminated area Contaminated area 250 257 2015-2016 n.a. 2.98 0.01 13.77 0.01 8.69 Source: 1) Ingelido A M et al. (2018). Biomonitoring of perfluorinated compounds in adults exposed to contaminated drinking water in the Veneto Region, Italy. Environment international, 110, pp.149-159. 2) Center for Disease Control and Prevention (2018). Fourth National Report on Human Exposure to Environmental Chemicals. Updated Tables Volume 1. 3) Frisbee S J et al. (2009). The C8 health project: design, methods, and participants. Environmental health perspectives, 117(12), 1873-82. Epidemiological studies on the health impacts of elevated levels of exposure are available from West Virginia and the Veneto Region. One West Virginia study323 found an increased risk of high cholesterol, while other studies found a higher risk of cancer (kidney and testicular) and hyperuricemia.324 A study from the Veneto Region found that residents suffered an increased risk of overall mortality due to exposure to PFAS from a nearby manufacturing plant.325 The valuation for health-related costs for population of affected communities focused on the all-cause mortality endpoint using the increased risk factor found in the Veneto Region study.326 The calculation relied on the 3% estimate from Sweden and assumed that the distribution of contaminated sites in Sweden is comparable to other European countries and that a similar share of the population is exposed at medium levels. In reality this is likely to be a low estimate - the prevalence of elevated exposure may be higher in countries with more manufacturers of PFAS-treated products and countries with higher population density. The assumption of 3% is however bolstered by findings from the United States where national drinking water monitoring data suggests that a higher share of the population (about 4.5%) is exposed to elevated levels of PFAS.327 323 Steenland K et al. (2009). Association of perfluorooctanoic acid and perfluorooctane sulfonate with serum lipids among adults living near a chemical plant. American journal of epidemiology, 170(10), pp.1268-1278. 324 Steenland K et al. (2010). Association of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) with uric acid among adults with elevated community exposure to PFOA. Environmental health perspectives, 118(2). p.229. and Barry V et al. (2013). Perfluorooctanoic acid (PFOA) exposures and incident cancers among adults living near a chemical plant. Environmental health perspectives, 121(11-12), p.1313. 325 Mastrantonio M et al. (2017). Drinking water contamination from perfluoroalkyl substances (PFAS): an ecological mortality study in the Veneto Region, Italy. The European Journal of Public Health. Feb 1;28(1):180-185. 326 Ibid. 327 The Groundwater Association (2017). PFAS Top 10 Facts. The cost of inaction linked to PFAS exposure 99 The calculation then assumes that the exposed adult population would face an elevated risk of mortality on the order of magnitude found in the Veneto region study. Using the baseline death rate reported by Eurostat328, the calculation suggests that more than 12,000 deaths annually could be attributed to elevated PFAS exposure in the EU with an estimated loss of EUR 43 billion. A similar calculation for only the Nordic countries329 suggests that almost 600 deaths annually are linked to PFAS, with a total value of EUR 2 billion. Table 22 presents a summary of the calculations, which indicate the potential scale of the health-related costs of elevated (medium) PFAS exposure in adults living in affected communities. Table 22: Monetised annual costs due to elevated risk of all-cause mortality for adults living close to PFAS contamination Nordic countries EEA countries "Exposed" population (3%) Annual deaths linked to PFAS Valuation of life lost 621 thousand 587-692 EUR 2.1-2.4 billion 12.5 million 11,745-13,843 EUR 41.1-48.5 billion Note: For more information on this calculation please refer to Annex 2. Pregnancy outcomes may also be affected in communities with elevated exposure. As seen in Table 23 low birth weight (weight less than 2,500 grams) is a health endpoint pointed out in several studies. The EFSA report notes that there is an overall tendency towards an inverse correlation between concentrations of PFOS/PFOA and birth weight330 whereas the five-year retrospective study conducted as part of the C8 Health Study in West Virginia331 found that PFOS, but not PFOA, to be associated with low birthweight. The prevalence of low birth weight in Europe is estimated to be 6.8%.332 Assuming that 3% of births take place in areas with elevated (medium) levels of exposure to PFAS, an estimated 3,544 births in EEA countries are low birth weight due to exposure to PFAS. In the Nordic countries, the prevalence of low birth weight is less (4.57%); in applying the same relative risk function, the analysis suggests that 271 births each year in Nordic countries are low birthweight due to medium level PFAS exposure. Low birth weight may be associated with a higher risk of developing diseases in adulthood such as cardiovascular disease, respiratory disease and diabetes.333 Low birthweight 328 Eurostat provides the rate of deaths per 100,000 for Europe. For the Nordic Countries, a weighted average was used. 329 Aside from the smaller population/exposed population, the key difference is the lower death rate for the Nordic countries, compared to the overall EU death rate. 330 EFSA CONTAM Panel (EFSA Panel on Contaminants in the Food Chain), Knutsen HK et al., 2018. Scientific opinion on the risk to human health related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. EFSA Journal 2018;16(12):5194, 128 pp. DOI: https://doi.org/10.2903/j.efsa.2018.5194 331 Stein C R et al. (2009). Serum levels of perfluorooctanoic acid and perfluorooctane sulfonate and pregnancy outcome. American journal of epidemiology, 170(7), pp.837-846. 332 OECD, 2016. Health status statistics. 333 Almond D and Currie J (2011). Killing me softly: The fetal origins hypothesis. Journal of economic perspectives, 25(3), 153-72. And Bharadwaj P et al. (2017). Birth Weight in the Long Run. Journal of Human Resources, 53(1), 189-231. 100 The cost of inaction linked to PFAS exposure is also associated with other important outcomes such as impaired cognitive development. For example, one study found that low birth weight was associated with a 25% lower likelihood of passing high school exit exams and a higher risk of unemployment at age 33 years.334 Other studies have found that birth weight is positively associated with earnings.335 For example, one study found that low birth weight was associated with lower income for men 30 years of age and for women between 50 and 60 years of age.336 The researchers found that part of this effect could be explained by a higher use of sick leave, which suggests a higher susceptibility to illness. This increased risk of illness could be relevant for 271 low birth weight persons born each year in the Nordic countries and for the estimated 3,544 low birth weight persons born each year in EEA countries. Table 23: Number of births per year that are low birth weight in areas close to PFAS contamination Nordic countries EEA countries Births in "exposed" areas Births of low birth weight linked with PFAS 8,843 births 271 births 156,344 births 3,544 births Note: For more information on this calculation please refer to Annex 2. 5.1.3 Background (low) exposure scenario PFAS are widely present in consumer goods as highlighted in Case Study 3 (see Section 3.3). Consumer goods can contribute to human exposure to PFAS through several pathways. First, it can enter humans through direct hand-to-mouth transfer through products such as food contact materials. Second, PFAS-treated consumer goods like cosmetic products may end up in wastewater that then affects the water supply. Some research suggests that children have an elevated exposure to PFAS and that this is due to frequent hand-to-mouth transfer and proximity to dust on the floor.337, 338 Young children may also have greater exposure to PFAS-treated carpets.339 While this conclusion of greater exposure for children is not supported by population-level bio- 334 Currie J and Hyson R (1999). Is the impact of health shocks cushioned by socioeconomic status? The case of low birthweight. American Economic Review, 89(2), 245-250. 335 Black S et al. (2007). From the cradle to the labor market? The effect of birth weight on adult outcomes. The Quarterly Journal of Economics, 122(1), 409-439 and Bharadwaj P et al. (2017). Birth Weight in the Long Run. Journal of Human Resources, 53(1), 189-231. 336 Bharadwaj P et al. (2017). Birth Weight in the Long Run. Journal of Human Resources, 53(1), 189-231.Table F. 337 Winkens K et al. (2017). Early life exposure to per- and polyfluoroalkyl substances (PFASs): A critical review. Emerging Contaminants. 3: 55-68. DOI: https://doi.org/10.1016/j.emcon.2017.05.001 338 Rappazzo K et al. (2017). Exposure to perfluorinated alkyl substances and health outcomes in children: a systematic review of the epidemiologic literature. International journal of environmental research and public health, 14(7). p. 691. Jun 27;14(7). pii: E691. DOI: https://doi.org/10.3390/ijerph14070691 339 Trudel, D., Horowitz, L., Wormuth, M., Scheringer, M., Cousins, I. T., & Hungerbhler, K. (2008). Estimating consumer exposure to PFOS and PFOA. Risk Analysis: An International Journal, 28(2), 251-269. The cost of inaction linked to PFAS exposure 101 monitoring data from the United States for three PFAS compounds (see Table 24) nevertheless a low level exposure to chemicals at critical periods of development such as infancy and childhood may have serious, irreversible impacts.340 Table 24: Serum concentration levels by age level (USA, 2013-2014) PFNA PFOA PFOS 3-5 years 6-11 years 12-19 years 20 years and older 0.764 0.809 0.599 0.685 2.00 3.38 1.89 4.15 1.66 3.54 1.98 5.22 Sample size 181 458 401 1,764 Source: CDC, 2018. Fourth National Report on Human Exposure to Environmental Chemicals. Updated tables, Volume 1. A review of studies on the health impacts of exposure to PFAS among children identifies four key health endpoints341, namely high cholesterol, depressed immunity (including lower vaccine response and higher risk of asthma), reduced renal function, and younger age at menarche. These endpoints are reflected in the overall review of endpoints presented in Table 25. It is challenging to monetise the increased risk of these health endpoints among children. It requires a life course view and a clear understanding of the complex relationships between the endpoints, other factors and the consequent impacts on development and well-being. For example, reduced renal function among children is associated with having delayed motor skills and language development, and trouble with concentration and selfesteem. Such issues could have cost implications in terms of more doctor visits, as well as poorer educational achievement. Among the four conditions mentioned above, depressed immunity was found to be the most amenable to quantification. As stated in Rappazzo et al., 2017, the studies of vaccine response were well done cohort study designs and despite the small number offer compelling evidence. The asthma studies are less consistent and include a broader range of study designs and quality.342 Several regulatory bodies highlight immunotoxicity as a likely health consequence of PFAS exposure (e.g. EFSA, ATSDR and US-EPA). However, a closer review of the evidence for specific endpoints presents a more mixed picture. For example, the US-EPA validation studies conclude that the relationship between prenatal and maternal exposure to PFOA and PFOS and infectious disease is limited. While some studies identified a relationship with common childhood infections such as otitis media, allergies, common colds and the flu, the evidence was not consistent. For example, the C8 340 As stated in Rappazzo et al. (2017): It is increasingly understood that exposure to environmental chemicals during sensitive windows has the potential to permanently alter a child's risk of future morbidity, even at doses that have little effect in adults. 341 Rappazzo K et al. (2017). Exposure to perfluorinated alkyl substances and health outcomes in children: a systematic review of the epidemiologic literature. International journal of environmental research and public health, 14(7). p.691. Jun 27;14(7). pii: E691. 342 Ibid. 102 The cost of inaction linked to PFAS exposure Science Panel concluded that there is no probable link between PFOA and common infections.343 If greater susceptibility to illness is indeed a health endpoint, PFAS exposure may then be linked to costs related to more doctor's visits, fewer days attending school and lower workplace engagement for caretakers. Increased sickness among children may be due to a wide range of factors other than PFAS exposure and thus it would be important to base any estimation on studies with a robust design. A Danish study explored the potential scale of the immunotoxicity endpoint. The study found that higher prenatal exposure to PFOS was associated with more days of fever among children ages 1 to 4 years.344 For this study a calculation was made assuming that the level of PFAS exposure and the level of risk of fever could be directly extrapolated to the Nordic and the EEA countries. The estimated number of days of fever that may be attributable to PFAS exposure are presented below. Table 25: Additional fever days among children ages 1-4 years due to PFAS exposure Nordic countries EEA countries "Exposed" children Fever days in exposed population - overall Fever days in exposed population - linked to PFAS exposure 45,229 212,576 83,742 784,794 3,688,533 1,453,059 Note: For more information on this calculation please refer to Annex 2. 5.1.4 Background exposure (low) scenario Discharges of PFAS from waste water treatment plants and landfills may lead to contamination of food and water, which are the two main sources of exposure to PFAS.345 Case Study 5 (section 3.5) notes that PFAS released from products in landfills can migrate directly into soil and groundwater. The discharge may include a high level of long-chain PFAS compounds such as PFOA even if their production has been restricted in recent years. Contaminated water can travel long distances and be taken up in agricultural produce and drinking water, potentially affecting all persons in Europe. Lastly, evidence suggests that wastewater treatment is not effective at fully removing PFAS within the general population. This study assessment found that exposure to PFAS through the pathway of waste water treatment plants and landfills may be most closely linked to background levels of PFAS in drinking water and food. Other factors may contribute to background levels of PFAS, for example, the disposal of consumer goods and exposure to dust. The blood serum concentration of European populations with background levels of exposure to PFAS is not well-known. Some biomonitoring studies have been conducted 343 Environment Protection Agency (2016). Health effects support document for perfluorooctanoic acid (PFOA). Document no: EPA822R16003 and Health effects support document for perfluorooctane sulphate (PFOS). Document No: EPA822R16002. 344 Dalsager L et al. (2016). Association between prenatal exposure to perfluorinated compounds and symptoms of infections at age 1-4years among 359 children in the Odense Child Cohort. Environment international, 96, pp.58-64. 345 Trudel D et al. (2008). Estimating Consumer Exposure to PFOS and PFOA. Risk Analysis: An International Journal 28.2: 251-269. The cost of inaction linked to PFAS exposure 103 among small populations as shown in Table 26. While the generalisability of these findings is limited due to the small sample sizes, the findings nevertheless provide insight into the levels of background PFAS exposure in different contexts. Table 26: Cross-country estimates of background PFAS levels in blood serum of adult populations Italy1 Norway2 Sweden3 Belgium4 Spain5 Germany6 Greece7 USA8 Year Sample size PFOA 2015-2016 2006 257 57 1.64 2.7 PFNA 0.58 0.55 PFDA 0.32 0.22 PFUnDA 0.18 0.14 PFDoDA 0.04 -- PFHxS 2.49 1.4 PFOS 5.84 12 1997-2000 66 1998, 2000 20 5.0 4.1 2006 48 3.4 -- -- -- -- -- -- -- -- -- -- -- -- 3.0 1.3 5.8 34.2 17.2 15.2 2005 356 13.7 -----5.7 2008 86 2.05 (male) 1.92 (female) 13.63 (male) 9.28 (female) 2009- 2014 2,000+ 3.07 0.675 0.185 --1.35 4.99 Source: 1) Ingelido A M et al. (2018). Biomonitoring of perfluorinated compounds in adults exposed to contaminated drinking water in the Veneto Region, Italy. Environment international, 110, pp.149-159. 2) Haug L S et al. (2009). Time Trends and the Influence of Age and Gender on Serum Concentrations of Perfluorinated Compounds in Archived Human Samples. Environ. Sci. Technol. 43:6, 2131- 2136. 3) Krrman A et al. (2004). Levels of perfluoroalkylated compounds in whole blood from Sweden. Organohalogen Compd. 66, 4008-4012. 4) Kannan K et al. (2004). Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ. Sci. Technol. 38, 4489-4495. 5) Ericson I et al. (2007). Perfluorinated chemicals in blood of residents in Catalonia (Spain) in relation to age and gender: a pilot study. Environ. Int. 33, 616-623. 6) Holzer J et al. (2008). Biomonitoring of perfluorinated compounds in children and adults exposed to perfluorooctanoate (PFOA) contaminated drinking water. Environ. Health Perspect. 116, 651-657. 7) Holzer J et al. (2008). Biomonitoring of perfluorinated compounds in children and adults exposed to perfluorooctanoate (PFOA) contaminated drinking water. Environ. Health Perspect. 116, 651-657. 8) Center for Disease Control and Prevention (2018). Fourth National Report on Human Exposure to Environmental Chemicals. Updated Tables Volume 1. In contrast to these small-scale biomonitoring studies, a large, nationally-representative population-based survey that includes biomonitoring is conducted every two years in the United States. The survey is known as the National Health and Nutrition Examination Survey (NHANES). One of the survey modules involves a health examination and the collection of a blood sample. The blood samples are analysed for a wide range of chemicals including PFAS. One analysis concluded that the levels of 104 The cost of inaction linked to PFAS exposure PFOS, PFOA and PFHxS in blood serum in the United States were comparable to several European countries.346 This conclusion supports the transferability of findings from studies conducted in the United States on the health impacts of background exposure to PFAS to the European context. Several analyses using the NHANES data suggest that PFAS exposure at background levels can increase the likelihood of cardiovascular disease. One study found a positive correlation with uric acid347 while another found an association with total cholesterol348, which are both risk factors for cardiovascular disease. Another study found an association between PFOA and risk of developing hypertension.349 The level of consensus was found to be highest for serum cholesterol. The epidemiological evidence relating elevated cholesterol and cardiovascular disease, however, is quite mixed. Another complicating factor is that studies investigating PFAS exposure do not assess the relative risk of cholesterol over the acceptable thresholds. Rather, they provide estimates for the elevation of cholesterol, which may remain below the acceptable threshold. Due to challenges related to the serum cholesterol endpoint, the quantitative assessment focused on hypertension. The risk relationship between PFAS and hypertension was based on findings from a study from the United States (see Table 27).350 Individuals with a systolic blood pressure greater than 140 mm Hg, a diastolic blood pressure greater than 90 mm Hg or a self-reported medical diagnosis were considered to have hypertension. Table 27: Risk of developing hypertension as a function of exposure to PFOA contamination Serum PFOA concentration Odds-ratios (95% confidence interval) Quartile 1 (<2.6 ng/ml) Quartile 2 (2.7-3.9 ng/ml) Quartile 3 (4.0-5.5 ng/ml) Quartile 4 (>5.6 ng/ml) 1 (Reference) 1.24 (0.89-1.74) 1.63 (1.20-2.20) 1.60 (1.15-2.22) Note: The analysis of hypertension is based on a sample of 2,208 adults (20 years and older) who provided a blood sample between 2003 and 2006 to the National Health and Nutrition Examination Survey. Source: Min et al., 2012. 346 Kato K et al. (2011). Trends in exposure to polyfluoroalkyl chemicals in the US population: 1999- 2008. Environmental science & technology, 45(19), pp.8037-8045. 347 Shankar A et al. (2012). Perfluorooctanoic acid and cardiovascular disease in US adults. Archives of Internal Medicine, 172(18), pp.1397-1403. 348 Nelson J et al. (2010). Exposure to polyfluoroalkyl chemicals and cholesterol, body weight, and insulin resistance in the general US population. Environmental health perspectives, 118(2), 197. 349Min J Y et al. (2012). Perfluorooctanoic acid exposure is associated with elevated homocysteine and hypertension in US adults. Occup Environ Med, 69(9):658-62. 350 Ibid. The cost of inaction linked to PFAS exposure 105 The analysis assumes that the entire adult population in Europe is exposed to background level exposure of PFOA through drinking water, and that about half have a level of exposure that is associated with a higher risk of developing hypertension. The latter assumption follows from a finding from the US study, where the risk for developing hypertension was elevated for individuals in the highest two quartiles of exposure. Several additional assumptions were made. For example, an estimated 6.1 million new cases of cardiovascular disease were diagnosed in the EU in 2015351, of which about half were assumed to be specifically related to hypertension.352 The assessment then considered the increased risk of mortality due to hypertension.353 These assumptions are presented in greater detail in Annex 2 and the findings are presented in Table 28. The analysis finds an estimated 12,655 to 41,417 cases of hypertension linked to PFAS exposure in the Nordic countries and about 153 to 500 deaths linked to hypertension and PFAS exposure. The estimated number of deaths that could be attributed to PFAS exposure in the EEA countries ranged from 3,066 to 10,035. Key uncertainties in developing these estimates is the underlying risk relationship between PFAS exposure and hypertension and the elevated risk of mortality associated with hypertension. Table 28: Monetised health impact (EUR) for background PFAS exposure leading to increased risk of developing hypertension Nordic countries EEA countries Population at elevated risk of hypertension Cases of hypertension linked to PFAS Deaths linked to hypertension and PFAS Valuation of life lost 10.3 million 12,655-41,417 153-500 EUR 687 million-2.2 billion 207.8 million 254,167-831,818 3,066-10,035 EUR 10.7 billion-35 billion Note: For more information on this calculation please refer to Annex 2. These figures do not capture all the costs associated with hypertension. Hypertension can also lead to other costs including health care costs, productivity lost and the cost of uncompensated care.354 Moreover, in general, these figures do not include utility costs, i.e. the benefits of being fit and well and enjoying life to the fullest. 351 European Heart Network (2017). European Cardiovascular Disease Statistics 2017. 352 In the same report it is stated that high systolic blood pressure contributes for about half of all cardiovascular diseases. 353 Zhou et al. (2018). Uncontrolled hypertension increases risk of all-cause and cardiovascular disease mortality in US adults: the NHANES III Linked Mortality Study. Scientific reports, 8(1), p.9418. 354 European Heart Network (2017). European Cardiovascular Disease Statistics 2017.An estimated 53 percent of the cost of cardiovascular disease (which includes hypertension) in the EU is accounted for by health care costs (EUR 111 billion), while 26 percent is due to productivity loss (EUR 54 billion) and the remainder is due to the provision of uncompensated care (EUR 45 billion). CVD can account for about 19 percent of all DALYs lost in the EU annually. 106 The cost of inaction linked to PFAS exposure 5.1.5 Summary of health-related costs of exposure to PFAS Table 29 presents an overview of the preliminary quantitative assessments of the health-related costs from PFAS exposure by the source. Table 29: Health-related costs (of exposure to PFAS) Exposure "Exposed" popu- level lation and source Health endpoint Nordic countries Population at risk Annual costs Occupa- Workers at chem- Kidney n.a. n.a. tional ical production cancer (high) plants or manu- facturing sites Elevated (medium) Communities near chemical plants, etc. with PFAS in drinking water Allcause mortality Low birth weight 621,000 EUR 2.1-2.4 billion 8,843 births 136 births of low weight Infection 45,000 children 84,000 additional days of fever Background (low) Adults in general population (exposed via consumer products, background levels) Hyper- 10.3 million ten-sion EUR 0.7-2.2 billion Totals Nordic countries EUR 2.8-4.6 billion All EEA countries Population at risk 84-273,000 Annual costs EUR 12.7-41.4 million 12.5 million EUR 41-49 billion 156,344 births 785,000 children 207.8 million 3,354 births of low weight 1,500,000n additional days of fever EUR 10.7-35 billion All EEA countries EUR 52-84 billion 5.2 Non-health costs of environmental contamination with PFAS Environment-related costs are considered here to cover the following elements: 1. Monitoring to assess PFAS contamination where it is suspected. 2. Provision of a temporary uncontaminated drinking water supply. 3. Upgrading of water treatment works and ongoing costs for maintenance and replacement and disposal of filters. 4. Excavation and treatment of soils. 5. Health assessments where contamination is found (health management costs, rather than costs of damage to health). 6. Impacts on biodiversity. Data collected during the study on elements 1 to 5 of this list are provided in Annex 3, Part 1, with summary estimates provided in this Section. The Annex also includes information on the Purchasing Power Parity (PPP)-adjusted exchange rates used and factors applied to account for inflation. The cost of inaction linked to PFAS exposure 107 5.2.1 Impacts on, and public aversion to, risks to the natural environment Direct assessment of impacts on biodiversity is not possible, given a lack of data on the stock at risk, exposure-response, and other elements of the analysis. However, it is noted that various ecological impacts may be associated with the release of PFAS, ranging from impacts to the endocrine and immune systems of animals to restrictions on human fishing activity (e.g in the cases of the contamination at Schiphol Airport and at Arnsberg in Germany, referred to above). For contamination of the River Rhine from Dusseldorf Airport, affected usages have been listed as being associated with the water works, local water rights including private wells, anglers, surfing club, and agricultural uses.355 For specific cases it may be possible to value elements such as lost fishing opportunity. Valuation data were generated for a specific case in the USA by Sunding as noted above, but this requires information on levels of fishing activity, the extent to which recommendations not to consume locally caught fish affects angling activity, the duration of any such recommendation, etc. and is beyond the scope of a generic study such as this. It is, however, important to note that Sunding's estimate based on willingness to pay to avoid fishing in PFOS contaminated surface waters for the one case of contamination arising from 3M's disposal of PFCs in Washington County, Minnesota amounted to the equivalent of over EUR 90 million for the period 2008-2040. This supports the view that loss of amenity associated with contamination of fish purely from the perspective of anglers is substantial. Section 3.2.3 reported on a UK study undertaken to inform the development of the REACH Restriction on D4 and D5 in wash-off personal care products. This found a willingness to pay of EUR 46 per year per person to reduce the risks associated with the PBT substance - D4, and EUR 40 per year per person to reduce the risks associated with the vPvB substance - D5. The results of the survey could be used to provide an estimate of the total willingness to pay to avoid contamination with PBT/vPvB substances including PFAS. Such a result could be useful in the context of evaluation of a REACH Restriction or Authorisation where the costs of alternatives are substantially less (e.g. an order of magnitude) than the WTP estimate, or where the WTP for specific beneficial properties associated with the substance under investigation are also provided for comparison (as was the case with the D4/D5 Restriction). For the purpose of the present study it seems better not to include the figures given the uncertainties that are associated with them, but to acknowledge firstly that economic estimates exclude ecological damage, and secondly that the costs linked to aversion to PBT and vPvB substances are likely to be substantial. The results of the UK study applied at a European scale suggest this aversion has an economic cost running into the billions of Euro. 355 Weber R (2016). Some lessons learned from PFOS/PFAS management in Germany. 108 The cost of inaction linked to PFAS exposure 5.2.2 Summary of the non-health unit costs associated with environmental contamination with PFAS Data from Annex 3, Part 1 on the non-health costs associated with environmental contamination with PFAS have been reviewed and collated to provide best estimates and associated ranges, see Table 30. The logic for defining best estimates and ranges is discussed below. Where possible, estimates are provided as cost per person to enable aggregation at a later stage. About half of the data that has been collected comes from the USA, raising questions about its transferability to Europe. Inspection of the data suggests that there is broadly as much variation within the European and US data sets, so both are included. Comparison of the US and European data could be taken to indicate that attitudes to risk aversion in the two regions are broadly similar (acknowledging the high level of variation in both the European and US datasets). An important factor constraining the ranges (even though many are broad) is the fact that when scaling up from information collected in the case studies, the analysis needs to adopt representative best estimates and associated ranges around those best estimates. The costs of individual schemes may well lie outside these ranges, but the analysis needs to deal with average costs, not the extremes from specific cases which by their nature would lead to an over-or under-estimation of costs. Table 30: Summary of cost data for non-health expenditures. For units, see second column Activity Unit Best estimate Range from studies Adopted range Monitoring Cost/sample Cost/case Health assessment (including biomonitoring) Provision of temporary uncontaminated supply Cost/person Total biomonitoring and health assessment per case where it is considered appropriate Cost/person Provision of a new pipeline Upgrading water treatment works (capital) Upgrading water treatment works (maintenance) Excavation and treatment of soils Cost/person Cost/person Cost/person Cost/kg PFAS Cost/case EUR 340 EUR 50,000 EUR 50 EUR 3.41 million EUR 278-402 EUR 4.0 thousands -6.1 million No range EUR 2.5 million - 4.3 million EUR 278-402 EUR 25 thousands -500 thousands EUR 5-95 (+/-90%) EUR 1 million -5 million No relevant data: Hoosick Falls information rejected as it does not appear to be for a true "temporary" solution (see text) EUR 800 EUR 37-5,000 EUR 500-1,500 EUR 300 EUR 8-2,200 EUR 18-600 EUR 19 EUR 8-30 EUR 8-26 EUR 280,000 EUR 5 million EUR 100 thousands -4.3 million EUR 100 thousands -3 billion EUR 100 thousands -1 million EUR 300 thousands -50 million With respect to monitoring, costs per sample concern the collection of individual samples of groundwater (etc.) and the analysis of those samples. Associated costs do not account The cost of inaction linked to PFAS exposure 109 for additional monitoring related activities, such as management of the monitoring regime. The reported costs per sample for monitoring were from a single study, but provided with a range indicating economies of scale as more monitoring was done. Relative to other parameters the range for monitoring is narrow, and the mid-point is adopted as the best estimate. A more important question for the analysis concerns how much monitoring would need to be carried out in total and in any country and with what frequency. This could vary from none or a few samples to several thousand. Figures expressed as average monitoring cost per case could therefore be more reliable. The range here is much larger, reflecting differences in the degree of contamination, and the extent of the population affected. Issues relating to the data will concern the activities included under "monitoring": in some cases these will cover only sampling and analysis, whilst in others they may include development of plans for public protection as well. Two sources of information concerning the costs of carrying out health assessments are considered356, WHO's evaluation of a health assessment scheme around the Veneto site of contamination, and biomonitoring costs incurred around Ronneby Airport in Sweden. Costs in both cases run into the millions of EUR. It is unknown how representative these cases are, but it is possible (given that this cost category has not been identified for other cases) that they are higher than similar costs incurred elsewhere if indeed such activities have been undertaken. For the Ronneby case there is no information on the scale of the biomonitoring undertaken, whereas WHO provide data on the number of individuals potentially affected in Veneto. Accordingly, the health assessment costs are taken from the Veneto case and cover surveillance of the population once significant exceedance has been observed. The Ronneby data are useful for indicating that where health assessment of some kind is considered appropriate, costs can run into millions of EUR. The best estimate is equal to the figure derived from the Veneto data (in this case, the number of people factored into the calculation of the average cost per person was the total exposed population, not the number of individuals undergoing testing of any kind). The extent of health assessment could vary significantly, theoretically from none at all to detailed and regular assessment of all exposed individuals. This variation could reflect national attitudes to pollution, public concern over a particular incident, the extent of exceedance of limit values, etc. A range of +/-90% is adopted in the costs analysis, reflecting possible variability. Two cases were identified for the costs of providing a temporary uncontaminated supply, from Hoosick Falls, New York State and Peterson, Colorado, both in the USA. Hoosick Falls was home to the Saint-Gobain Performance Plastics and Honeywell Manufacturing Plant, whilst Peterson contains a US Air Force Base. For Hoosick Falls a `temporary filtration system' was installed. Associated costs also include investigation of alternative water sources. The Hoosick Falls estimate of EUR 7.4 million for provision of a temporary water filtration system is of a similar magnitude to the costs reported elsewhere for a permanent system. This suggests that the responsible authorities are seeking an alternative source with no contamination but that this will take several years to 356 The health assessment costs are considered in this section, rather than Section 5.1 as they relate to management of problems of PFAS contamination rather than health impacts. 110 The cost of inaction linked to PFAS exposure come online. In the meantime they have introduced a system that is similar to the permanent solutions elsewhere. Bottled water has been supplied to residents in both areas. A cost of EUR 81,000 equivalent is cited for Peterson, but no cost data for bottled water have been identified for Hoosick Falls. Considering estimation of the costs of provision of temporary clean water supplies elsewhere, it is noted that costs will vary substantially according to the following factors: the number of people served by a scheme; the duration over which alternative supplies are required; and the form of the temporary measures that are put in place (e.g. additional filtration, provision of bottled water, provision of water tankers, temporary piping). On the basis that PFAS contamination will not be resolved quickly, the duration over which alternative supplies are required seems likely to be in the order of months or years rather than days or weeks. This in turn starts to rule out very prolonged use of some of the quick fixes such as provision of bottled or tanked water. Four of the incidents for which cost data were obtained provide estimates indicating the costs of providing permanent new pipelines, these being the cases for Jersey, Kallinge, Stadtwerke Rastatt and Veneto. There are two orders of magnitude difference in the reported costs which show a strong dependence on the number of people presumed to be affected. Factors affecting costs include: whether a new connection is made to a single point that goes on to serve a larger area, as assumed for Rastatt, or to individual buildings as was the case for Jersey; the number of people affected; the distance over which new pipework is required; and complications in laying pipes associated with geology and local infrastructure The costs in Jersey are significantly higher than the other two locations, reflecting the small number of houses affected and the need to connect to individual homes rather than a single point. The likely range for an average cost per person is taken from the lower bound to the upper bound of costs excluding Jersey (EUR 100-EUR 1,500/person) with the best estimate taken as the mid-point of this range (EUR 800). For upgrading of water treatment plant the cost range is again broad, EUR 8 to EUR 2,200 when normalised against population (the "high" estimate is notably the supposed cost of the temporary solution at Hoosick Falls). Excluding the two highest and lowest values gives a reduced range of EUR 18 to EUR 940/person, with remaining values spread rather evenly over this range. Taking the mean of the 9 data points excluding the highest and lowest gives a best estimate of EUR 300/person. For maintenance costs, there are four estimates for which normalisation against population has been performed, from EUR 8/person/year to EUR 30/person/year. The range is adopted from these studies and the best estimate (EUR 19) is taken as the midpoint. The cost of inaction linked to PFAS exposure 111 For excavation of soils the range is again broad, with costs per kg of PFAS of EUR 100,000, EUR 200,000-280,000 and EUR 4.3 million. It is unclear whether all of these costs can be attributed to PFAS or to other contaminants. Soil remediation costs can also be expressed on a per case basis. The range is again very broad, EUR 100,000 to EUR 3 billion, the very high upper bound referring to possible costs at Rastatt in Germany.357 Adopting any value within this range is prone to very high uncertainty, though for the purpose of illustration, a best estimate of EUR 5 million/case is taken, with a range for the main cost of EUR 300,000 to EUR 50 million. Definition of the upper bound cost for soil remediation is extremely difficult, given the extreme variability in the cost data identified (see Annex 3). Most cases identified had costs ranging from EUR 1 million to EUR 10 million, but there are several that are substantially higher (Schiphol at EUR30-40 million, Dusseldorf Airport where costs are estimated at up to EUR 100 million, and Baden-Wurttemberg where costs are estimated between EUR 1 and 3 billion). A number of other costs have also been identified for individual case studies, such as lost opportunity from closure of a borehole (Buncefield, UK), fees and capital works at specific sites not specifically involved in water treatment (Jersey) and risk analysis and project management (Uppsala). In each case, the total costs under this "other" category are substantial, covering a range of EUR 320,000 to EUR 6.3 million. Whilst no attempt is made to extrapolate these figures, given that they are only mentioned for single locations, they provide further evidence that the costs associated with remediation of PFAS contamination are large. 5.2.3 Aggregating the costs of environmental contamination with PFAS A first step in aggregation is to simply combine the results for each European country as provided in Annex 3 part 1 (see Table 31 below). The results given do not represent an estimate of total damage for any country, as such a total has not been estimated for any country. The country for which data appear most complete is Sweden, where significant contamination has been found at 20% (7) out of the country's 35 airports (military and civilian combined). This figure of 20% is carried forward to the analysis that follows. The country for which highest costs are estimated is Germany, where the total is almost entirely due to soil remediation in Baden-Wurttemberg, which is understood to follow the use of waste material as a soil treatment on agricultural land.358 It is further understood that the remediation of the soils in question has not been performed, so the cost estimate must be considered theoretical. The results demonstrate a variation in the estimated costs by around 2 orders of magnitude. 357 http://greensciencepolicy.org/wp-content/uploads/2016/09/Rolland-Weber-PFOS-PFAS-German-activities-Final.pdf 358 Reuning A, Landschaft mit Gift, Deutschland Funk, 23.04.2017. Accessed 10.11.2018. 112 The cost of inaction linked to PFAS exposure Table 31: Summation of the costs identified in the available literature, EUR millions by country. Figures in brackets indicate the number of cases or plant for which data were collected Total costs DK DE IT NL NO SE UK Monitoring Upgrading treatment works Install new pipelines 20 year maintenance cost of water treatment works discounted at 4% Soil remediation costs Biomonitoring EUR 15 EUR 7.8 (2) EUR 104 (2) EUR 1.81 (1) EUR 11 (1) EUR 2.1 (1) EUR 3,112 (2) EUR 35 (1) Other quantified costs Total quantified cost Sites affected EUR 15 Copenhagen (AP) EUR 3,236 Dusseldorf (fire), Rastatt (waste) EUR 2.1 Veneto industry EUR 35 Schiphol (AP) EUR 0.50 (2) EUR 4.1 (1) EUR 4.6 Oslo Fjord, Tyri-fjorden EUR 2.6 (7) EUR 5.3 (5) EUR 11 (2) EUR 1.6 (1) EUR 1.0 (1) EUR 13 (1) EUR 2.5 (3) EUR 2.6 (1) EUR 0.38 (1) EUR 26 Arlanda, Bromma, Kallinger, Kiruna, Ronneby, Umea, Uppsaala (all APs) EUR 7.1 (1) EUR 12 (2) EUR 33 Buncefield (fire), Jersey (AP) The results shown in Table 31 are incomplete because they omit most European countries and are based on limited knowledge of contamination across the continent. As such they describe the absolute minimum for addressing the PFAS problem to the extent that they have been incurred (noting the discussion above concerning the Baden-Wurttemberg case where soil remediation is still to take place). As a minimum estimate the figures are clearly not reliable. In the absence of a European wide systematic screening programme it is unlikely that all cases of contamination have been identified. 5.2.4 Number of sites releasing PFAS Aggregation of the cost data to provide some estimate of damage at Nordic and EU scales requires additional data, provided in Annex 3, Part 2, covering: population water consumption quantity of water supplied from groundwater and surface water number of wastewater treatment plant (WWTP) number of plant or sources providing drinking water number of PFAS producers The cost of inaction linked to PFAS exposure 113  number of companies potentially using PFAS number of airports number of landfill and incineration sites. Information covers the EU28, Norway and Iceland to the extent that data are available. Data for the USA are also included for reference, given that much of the information used in this report is of US origin. A summary of the number of sites releasing or potentially releasing PFAS is provided in Table 32 below, drawing on the information in Annex 3, Part 2. Table 32: Number of installations working in the sectors that may use or emit PFAS for the EEA. Figures in brackets represent businesses with more than 250 employees Sector Activity Total Waste water treatment plant Drinking water treatment Aviation Other fire control Waste Manufacturing industry T1 (primary treatment) T2 (secondary treatment) T3 (tertiary treatment) Of which T3N (T3 + nitrogen removal) Of which T3P (T3 + phosphorus removal) Large Small Very small Main passenger airports Medium passenger airports Small airports Military airbases Fire stations Site emergency services Hazardous waste landfill Non-hazardous waste landfill Large incineration (as energy from waste) PFAS manufacturers1 Textiles Leather Carpet Paper Paints and varnishes Cleaning products Cosmetics and personal care Electronics Photography films Metal treatments Car washes Mining Plastic, resins, rubbers 7,279 24,316 19,716 11,502 10,436 Thousands Ten thousands Hundred thousands 318 137 no data 239 84,099 no data 365 3,801 808 12-20 61,685 (262) 37,120 (159) no data 19,477 (488) 4,027 (104) (178) no data no data no data 151,455 (163) 79,000 no data (340) Note: 1) It has not proven possible to identify the European PFAS manufacturers with confidence. From data collected, it is assumed that there are between 12 and 20 sites, best estimate 16, distributed as follows: Belgium (2), Czechia (1), France (3), Germany (3), Italy (2), Netherlands (1), Poland (1), UK (3). 114 The cost of inaction linked to PFAS exposure Not all of these enterprises, particularly in manufacturing, produce, use or emit PFAS. However, it can be concluded that the total number of sites emitting PFAS in some quantity could be in the order of 100,000 or more for the EEA. Information presented in Annex 3 demonstrates that these activities are not concentrated in a few countries but are spread throughout the region. 5.2.5 Extrapolation of costs, by country Noting the uncertainties identified elsewhere in this report, a precise quantification of costs associated with PFAS contamination is not possible. Estimation of damage is, however, still an important exercise as it provides opportunity to describe the likely magnitude of economic damage. With this in mind, a scenario based assessment has been carried out. The following elements are estimated, with results in this section provided for each of the Nordic countries and the EU28+Switzerland combined: costs of a screening programme costs of monitoring costs of water treatment costs of soil remediation costs of health assessment studies. A full breakdown of results by country is provided in Annex 3, Part 3. A key input to the assessment is based around the Swedish National Food Agency's limit value for PFAS in drinking water of 0.09 g/l (90 ng/l) and techniques considered appropriate to treat contaminated water to meet that limit, and specifically the proportion of people exposed to levels above the limit value. Swedish data359 indicates that this applies to between 2% and 3% of the Swedish population. The higher figure is selected here as the analyses have focused on 7 PFAS compounds rather than the 11 currently covered by the limit value: increasing the number of compounds will clearly increase the reported concentration and make exceedance of the limit more likely. A range of 1% to 5% is applied around this estimate. The assessment identifies the number of sites where significant contamination (i.e. in excess of the Swedish limit value) has been found and applies the proportion of the population with significant exposure indicated in the Swedish data. 359 Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). The cost of inaction linked to PFAS exposure 115 Results are aggregated in the following tables to provide estimates for: the costs of a basic screening programme (Table 33); costs for monitoring at sites where significant PFAS contamination has been found (Table 34); costs for improvements to water treatment works to reduce exposure to PFAS above possible limits (Table 35); costs for soil remediation (Table 36); costs for health assessments when significant contamination is found (Table 37); and total of the above (Table 38). The estimates are based on a number of assumptions which are summarised in the text below each table. Definition of low, best and high estimates is not straightforward, given limited data. In several cases the lower bound is based on data for Sweden because it provides the most complete infomation available for any country. Since this information only relates to contamination associated with the use of AFFFs at airfields it is likely to provide a lower bound: as other data in this report show, there are numerous other sources of PFAS contamination present in Europe including the manufacturing processes, the use and the disposal of contaminated waste materials. Ranges are provided along with best estimates based on the data in Table 30. Estimated costs for a basic monitoring program are shown in Table 33. The assumptions used for the best estimate, and the low and high bounds, are given below the table. The best estimate indicates a cost in the order of EUR 14 million, in a range of EUR 2.8 to EUR 54 million. Table 33: Estimated costs for a basic screening programme to assess PFAS levels N facilities for best estimate Best estimate, EUR millions Low, EUR millions High, EUR millions Denmark Finland Iceland Norway Sweden Other EU28+CH Total 78 184 7 179 411 12,914 13,772 EUR 0.08 EUR 0.19 EUR 0.01 EUR 0.18 EUR 0.42 EUR 13.17 EUR 14.05 EUR 0.02 EUR 0.06 EUR 0.00 EUR 0.04 EUR 0.12 EUR 2.52 EUR 2.77 EUR 0.31 EUR 0.65 EUR 0.02 EUR 0.68 EUR 1.50 EUR 50.98 EUR 54.13 Best estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 3 samples, using best estimate of cost/sample for monitoring; 2. 5% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples; 3. Best estimate of costs adopted. 116 The cost of inaction linked to PFAS exposure Low estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 3 samples, using low cost/sample for monitoring. 2. 1% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples. 3. Low estimate of costs adopted. High estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 3 samples, using low cost/sample for monitoring. 2. 10% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples. 3. High estimate of costs adopted. A number of factors could influence the costs estimated for such a screening programme, including the number of samples taken at each point, whether sampling is carried out once only or repeatedly over time, how the programme is organised, whether it is specific to PFAS or whether the opportunity is taken to investigate the presence of other contaminants, and so on. Estimated costs for monitoring at contaminated sites are shown in Table 34. Table 34: Estimated costs for monitoring at sites where significant PFAS contamination has been found N facilities for best estimate Best estimate, airfields and PFASmanufacturing only, EURmillions Best estimate, all source categories included, EURmillions Low, EURmillions High, EURmillions Denmark Finland Iceland Norway Sweden Other EU28+CH Total 8 22 1 19 48 1,327 1,426 EUR 0.03 EUR 0.47 EUR 0.04 EUR 0.20 EUR 0.76 EUR 5.46 EUR 6.96 EUR 0.40 EUR 1.11 EUR 0.05 EUR 0.97 EUR 2.40 EUR 66.36 EUR 71.28 EUR 0.05 EUR 0.19 EUR 0.01 EUR 0.13 EUR 0.36 EUR 7.57 EUR 8.30 EUR 7.98 EUR 20.81 EUR 0.86 EUR 18.87 EUR 44.97 EUR 1,322.66 EUR 1,416.13 Best estimate assumptions 1. Assumed 20% of airports and airfields (as in Sweden) and PFAS manufacturing sites require monitoring programme, using best estimate cost/case for monitoring. 2. 0.5% of other facilities require monitoring. 3. Best estimate of costs adopted. The cost of inaction linked to PFAS exposure 117 Low estimate assumptions 1. Assumed 10% of airports and PFAS manufacturing sites require monitoring programme, using low estimate cost/case for monitoring. 2. 0.1% of other facilities require monitoring. 3. Low estimate of costs adopted. High estimate assumptions 1. Assumed 30% of airports and PFAS manufacturing sites require monitoring programme, using high estimate cost/case for monitoring. 2. 1% of other facilities require monitoring. 3. High estimate of costs adopted. Table 34 separates the best estimate of cost associated with airfields and PFAS manufacturing from the best estimate taking into account all source categories. The total for airfields and PFAS manufacturing is only 10% of the total for the "all source" best estimate, but broadly in line with the low estimate. Summing the data from the literature review (Annex 3, Part 1) for existing monitoring at European sites where significant contamination has been found gives a cost of EUR 10.9 million, also of a similar order of magnitude to the calculated lower bound, though dominated by the Baden-Wurttemberg case. The likelihood of PFAS contamination being restricted to airfields and the small number of PFAS manufacturers present in Europe is unrealistic, hence a figure towards the best estimate does not seem unreasonable. Under the high estimate, results are almost totally dominated by sources others than airfield and PFAS manufacture because of the large number of potential sites (even though only 1% are considered). This position seems highly unlikely based on those countries where extensive monitoring has already been undertaken. However, the potential for more extensive contamination than is currently recognised exists, as shown by the case of Rastatt in Baden-Wurttemberg, Germany, where widespread contamination appears to have arisen from the spreading of waste paper materials on agricultural land, ironically for the purpose of soil improvement.360 The quantification of the costs associated with improvements to water treatment is carried out by consideration of the fraction of the Swedish population exposed to levels of PFAS in excess of the Swedish limit value. In line with the estimate given above, the best estimate is taken as 3% of the population. Broadly similar results were obtained for the USA, providing some level of verification on the order of magnitude of the estimate (though accepting the potential for coincidence, given the limited evidence base). The analysis then uses the estimates of cost/person in affected areas given in Table 35. Costs include both the upgrading of water treatment works and operation and maintenance over a 20 year period. The period of 20 years is selected to reflect that in some cases, perhaps most, advanced treatment of water supplies will be needed for many years to come, running to many decades unless remediation actions are possible and implemented. 360 https://www.faz.net/aktuell/wissen/baden-wuerttemberg-chemische-abfaelle-auf-dem-acker-14419295.html 118 The cost of inaction linked to PFAS exposure In some cases, however, the need to use advanced water treatment may be reduced through connection to alternative uncontaminated supplies. Table 35: Estimated costs for improvements to water treatment works to reduce exposure to PFAS above possible limits. Population affected, best estimate Best estimate, EURmillions Low, EUR millions High, EUR millions Denmark Finland Iceland Norway Sweden Other EU28+CH Total 170,000 160,000 10,000 150,000 290,000 15,000,000 16,000,000 EUR 97 EUR 93 EUR 6 EUR 88 EUR 166 EUR 8,456 EUR 8,906 EUR 7 EUR 7 EUR 0 EUR 7 EUR 13 EUR 650 EUR 684 EUR 274 EUR 265 EUR 16 EUR 250 EUR 472 EUR 23,982 EUR 25,258 Best estimate assumptions 1. Assumed 3% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best es- timate. 3. Assume best estimate cost per case for water treatment. 4. Assume 4% discount rate on future maintenance costs. Low estimate assumptions 1. Assumed 1% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best es- timate. 3. Assume low estimate cost per case for water treatment. 4. Assume 4% discount rate on future maintenance costs. High estimate assumptions 1. Assumed 5% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best es- timate. 3. Assume high estimate cost per case for water treatment. 4. Assume 4% discount rate on future maintenance costs. In comparison, the results from the identified literature gives a figure of EUR 136 million for improvements to water quality, only 20% of the lower bound calculated here. However, the costs identified in the literature are in many cases based on situations where measures are yet to be taken. Also, further costs are not considered in this estimate, for example, improvement of water treatment, the need to construct new water pipelines The cost of inaction linked to PFAS exposure 119 to bring in clean water from other areas, or the costs of shutting down existing wells and losing the water resource that they provide. Therefore, the estimated lower bound may not be unrealistic, and it is possible that the costs are substantially higher. The results for individual countries will be more uncertain than the total estimate for the EEA countries. For example relative to Sweden the levels of contamination found in Denmark are much lower. This is likely linked to the distance between PFAS use and the location of the groundwater sources that dominate water supply. Table A3.9 in Annex 3 part 2 shows that 99% of drinking water in Denmark comes from ground rather than surface sources. This is a much higher percentage than in other countries. Extrapolation based on the Danish situation would therefore not account for the costs incurred already across Europe. Estimated costs of soil remediation are given in Table 36, again with details on assumptions for the best, low and high estimates provided below the table. Table 36: Estimated costs for soil remediation Number of sites affected for the best estimate Best estimate, EUR millions, airfields and PFAS manufacturing only, EUR millions Best estimate, EUR millions Denmark Finland Iceland Norway Sweden Other EU28+CH Total 8 22 1 19 48 1,327 1,426 EUR 3 EUR 47 EUR 4 EUR 20 EUR 76 EUR 546 EUR 696 EUR 40 EUR 111 EUR 5 EUR 97 EUR 240 EUR 6,636 EUR 7,128 Low, EUR millions High, EUR millions EUR 0.5 EUR 2.2 EUR 0.1 EUR 1.6 EUR 4.3 EUR 91 EUR 100 EUR 798 EUR 2,081 EUR 86 EUR 1,887 EUR 4,497 EUR 132,266 EUR 141,613 Best estimate assumptions 1. Assumed 20% of airports and PFAS manufacturing sites require remediation. 2. 0.5% of other facilities require remediation. 3. Best estimate of costs adopted. Low estimate assumptions 1. Assumed 10% of airports and PFAS manufacturing sites require remediation. 2. 0.1% of other facilities require remediation. 3. Low estimate of costs adopted. High estimate assumptions 1. Assumed 30% of airports and PFAS manufacturing sites require remediation. 2. 1% of other facilities require remediation. 3. High estimate of costs adopted. Of the estimates presented, soil remediation costs are likely to be the most uncertain, which is reflected in the range of the costs. The costs identified in the literature review 120 The cost of inaction linked to PFAS exposure for the Baden-Wurttemberg case alone are between EUR 1-3 billion361, substantially more than the low estimate and almost half the best estimate. However, that cost is an estimate, and it is possible that more cost-effective solutions will be found. As noted above, Baden Wurttemberg is not the only location where high costs have been reported, with Schiphol and Dusseldorf Airports also reporting high remedation costs. It is necessary to ask to what extent the costs of additional water treatment and soil remediation should be considered additive. It could be argued that with soil remediation carried out, there should be no need for additional water treatment, and vice-versa. However, this would ignore the time taken to carry out soil remediation, time during which water would need to be treated, or brought in from outside areas, to avoid excess exposure of the population. A failure to clean the soils would mean that water treatment would need to persist into the far future, rather than the 20 year period assumed here. Whilst there is some overlap it is clearly not a simple binary choice to treat water or remediate the soil, and both will often be necessary. A further factor to consider is that the longer soils are left contaminated, the more the PFAS will spread, potentially making clean-up more difficult, far more extensive and more expensive. Estimated costs for health assessments are provided in Table 37. Table 37: Estimated costs for health assessments when significant contamination is found Population affected, best estimate Best estimate, EUR millions Low, EUR millions High, EUR millions Denmark Finland Iceland Norway Sweden Other EU28+CH Total 170,000 160,000 10,000 150,000 190,000 14,000,000 15,000,000 EUR 8.5 EUR 8.2 EUR 0.5 EUR 7.7 EUR 15 EUR 744 EUR 783 EUR 0.28 EUR 0.27 EUR 0.02 EUR 0.26 EUR 0.49 EUR 24.79 EUR 26.11 EUR 27 EUR 26 EUR 1.6 EUR 25 EUR 46 EUR 2,355 EUR 2,480 Best estimate assumptions 1. Assumed 3% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume best estimate cost per case for remediation. Low estimate assumptions 1. Assumed 1% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume low estimate cost per case for remediation. 361 Cost data for this case are taken from http://greensciencepolicy.org/wp-content/uploads/2016/09/Rolland-Weber-PFOSPFAS-German-activities-Final.pdf. A full, detailed account of the costing has not been identified. However, further accounts (e.g. https://www.faz.net/aktuell/wissen/baden-wuerttemberg-chemische-abfaelle-auf-dem-acker-14419295.html) refer to this case as the largest environmental scandal in Germany in terms of the area affected (currently arounf 400 hecatres). The cost of inaction linked to PFAS exposure 121 High estimate assumptions 1. Assumed 5% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume high estimate cost per case for remediation. The need to undertake health assessment will vary from location to location. The cost of biomonitoring in Ronneby, Sweden averaged over EUR 0.5 million per year for a 3 year period. The WHO report health assessment costs for the Veneto region of EUR 4.6 million. On this basis, only a very small number of cases would be needed to reach the low estimate made here, so again, a figure between the low and best estimates is easily feasible. 5.2.6 Summary of environment-related costs of exposure to PFAS Table 38 shows total costs from the preceding tables covering initial environmental screening, monitoring where contamination is found, water treatment, soil remediation and health assessment, with a more detailed breakdown for the Nordic countries in Table 39. Table 38: Aggregated costs covering environmental screening, monitoring where contamination is found, water treatment, soil remediation and health assessment Best estimate, EUR millions Low, EUR millions High, EUR millions Denmark Finland Iceland Norway Sweden Other EU28+CH Total EUR 145 EUR 214 EUR 12 EUR 194 EUR 423 EUR 15,915 EUR 16,902 EUR 8 EUR 10 EUR 1 EUR 9 EUR 18 EUR 776 EUR 821 EUR 1,106 EUR 2,393 EUR 105 EUR 2,181 EUR 5,061 EUR 159,976 EUR 170,821 Table 39: Detailed breakdown for non-health environmental costs per Nordic Countries N people affected (3%) Screening and monitoring Health assessment Upgrade treatment works and maintenance Soil remediation Total (EURmillions) Denmark Finland Iceland Norway Sweden Nordic total 169,791 164,153 10,102 154,995 292,421 791,462 EUR 0.07-8.3 EUR 0.25-22 EUR 0.01-0.9 EUR 0.17-20 EUR 0.48-47 EUR 0.28-27 EUR 0.27-26 EUR 0.02-1.6 EUR 0.26 -25 EUR 0.49-46 EUR 7.4-274 EUR 7.2-265 EUR 0.4-1.6 EUR 6.8-250 EUR 13-472 EUR 0-798 EUR 2.2- 2,081 EUR 0.1-86 EUR 1.6-1,887 EUR 4.3-4,497 EUR 8-1,106 EUR 10-2,393 EUR 1-105 EUR 9-2,181 EUR 18-5,061 EUR 46- 10,846 122 The cost of inaction linked to PFAS exposure Parallel calculations for all 31 EEA Member Countries and Switzerland arrive at a range of costs for environmental remediation totalling EUR 821 million to 170 billion. The uncertainties associated with these estimates are clearly high, reflecting the level of variation in the unit cost estimates. For example, remediation of just the one case of soil contamination in Germany at Baden-Wurttemberg has been estimated at EUR 3 billion, though this work has yet to be undertaken. Uncertainties have been reviewed in the context of each of the tables presented. The lower and upper bounds should be considered illustrative because of the limited information available. Further analysis may be able to limit the range beyond that estimated here. However, based on the information from the literature review there is a firm basis for concluding that the lower bound estimates would be exceeded. A best estimate in the order of EUR 10-20 billion is certainly feasible. Significantly higher costs are feasible if several cases similar to the contamination at Baden-Wurttemberg are identified. It is noted that results for individual countries will be more uncertain than results for Europe as a whole. It is, for example, understood that the contamination levels in Denmark are lower than Sweden. The situation in other countries may vary, with different levels of contamination. The data available from Sweden only link contamination to airports, whereas problems elsewhere have been identified in relation to other sources such as PFAS manufacture and use (Veneto), fires (Dusseldorf and Buncefield), and spreading of waste paper materials on agricultural land as a soil improver (Rastatt). A further source of uncertainty concerns those elements of the analysis that have not been quantified. Notable amongst these are potential ecological effects, especially given the extreme persistence of PFAS. From an economic perspective the material reviewed above demonstrates public willingness to pay to avoid exposure to persistent and bioaccumulative substances, even without detailed assessment of impacts. Aggregated across the European population the willingness to pay based on information in the restriction proposal for D4 and D5 would be substantial in the order of billions of Euro and broadly comparable with the best estimate made here for monitoring and clean-up. The significant uncertainties in those estimates are recognised. Other elements that have not been quantified include the costs of providing new pipelines to provide access to uncontaminated water supplies and various administrative costs. The former, in particular, could be substantial based on evidence from a limited number of sites (e.g. Jersey, UK), certainly adding weight to the overall best estimate. The cost of inaction linked to PFAS exposure 123 124 The cost of inaction linked to PFAS exposure 6. Conclusions 6.1 Findings and discussion This study investigates the socioeconomic costs that may result from impacts on human health and the environment from the use of PFAS. Better awareness of the costs and long-term problems associated with PFAS exposure will assist authorities, policy-makers and the general public to consider more effective and efficient risk management. The production of PFAS, manufacture and use of PFAS-containing products, and end-of-life disposal of PFAS have resulted in widespread environmental contamination and human exposure. PFAS have been found in the environment all around the world and almost everyone living in a developed country has one or more PFAS in his/her body. Because of the extreme persistence of PFAS in the environment, this contamination will remain on the planet for hundreds if not thousands of years. Human and environmental exposure will continue, and efforts to mitigate this exposure will lead to significant socioeconomic costs - costs largely shouldered by governments and taxpayers. The focus of this study is on costs of inaction with respect to regulation of PFAS in the countries comprising the European Economic Area (EEA), i.e. the costs that society will have to pay in the future if action is not taken to limit PFAS-emissions today. The impact pathway concept provides an overall analytical framework for the study. Five case studies, based on literature reviews, following the life-cycle of PFAS are presented to illustrate the links between production, use and disposal of PFAS, health and environmental exposures, and impacts and their economic valuation. The pathway concept provides a template for assessment of each source, enabling the analyst to consider which impacts are relevant. The first three case studies cover the activities that account for a large proportion of the PFAS released into the environment: their production, their use in product manufacturing, and the use phase of PFAS-containing products. The industrial facilities producing the fluorochemicals and fluoropolymers, while relatively limited in number, are significant emitters of PFAS into the air, soil and waterways. Case Study 1 estimates that up to 20 facilities actively produce fluorochemicals in Europe, that these facilities are significant sources of PFAS released to the environment, and that exposure to workers at these plants is high. Other industrial activities with the potential to release PFAS to the environment take place throughout Europe, including the Nordic region. Case Study 2 considers the manufacture and commercial use of PFAS-containing products, including textile and leather manufacturing; metal plating, including chromium plating; paper and paper product manufacturing; paints and varnishes; cleaning products; plastics, resins and rubbers; and car wash establishments. Releases of PFAS occur via the air or effluent entering sewerage and wastewater treatment plants, before discharge into waterways. The cost of inaction linked to PFAS exposure 125 The third major source responsible for PFAS released to the environment, which is the focus of Case Study 3, is the widespread use of aqueous film-forming foams (AFFFs). The AFFFs are used to extinguish fires in emergencies or during training, especially around airports and military bases. Where the AFFFs have migrated to groundwater and other sources of drinking water, nearby communities have been affected by elevated levels of PFAS in their drinking water. It is noted that other uses of AFFFs for firefighting, especially at major industrial facilities, may also be a significant source, but one that has so far received little attention. Case Study 4 and 5 considers the use and the end-of-life phase of consumer products, which account for the remaining releases and direct sources of exposure to PFAS. Case Study 4 looks at PFAS-treated carpets, PFAS-treated food contact materials and cosmetics as examples of how a product's use is likely to lead to human exposure. Possible exposure occurs through ingestion and dermal absorption, or through releases to the environment when the product is washed off or laundered, entering sewers, treatment plants, and eventually waterways. The availability of suitable non-fluorinated alternatives makes the use of PFAS in many of these products unnecessary. Case Study 5 looks at end-of-life impacts of PFAS-treated products. Waste incineration may destroy PFAS in products if 1000 C operating temperatures are reached, but such temperatures are not typical of most incineration capacity (the EU Industrial Emissions Directive, for example, requires a temperature of 850 C for municipal waste incineration). If landfilled, the PFAS will remain even after the product's core materials break down. The compounds will eventually migrate into liquids in the landfill, then into leachate collection systems or directly into the natural environment. They may then enter drinking water supplies, be taken up by edible plants and bioaccumulate in the food chain. 6.1.1 Health-related costs to society To calculate health-related costs to society, the study looked for consensus regarding health endpoints affected by exposure to PFAS. Some agreement has emerged concerning liver damage, increased serum cholesterol levels (related to hypertension), decreased immune response, increased risk of thyroid disease, decreased fertility, pregnancy-induced hypertension/pre-eclampsia, lower birth weight, and testicular and kidney cancer. The methodology draws upon risk relationships developed in the course of specific epidemiological studies for populations exposed to PFAS at different levels. Workers exposed to PFAS in the workplace were used to exemplify a high level of exposure. Communities affected by PFAS, e.g. because of proximity to manufacturing sites or sites where fluorinated AFFFs were used were assumed to have been exposed at a medium level; this level of exposure was assumed to have been experienced by 3% of the European population. The general population was considered to have experienced exposure at low (background) levels. 126 The cost of inaction linked to PFAS exposure Table 40 provides an overview of the estimated annual costs for just a few health endpoints where risk ratios were available for affected populations. Despite the high level of uncertainty and the assumptions underlying the calculations, the findings suggest that the health-related costs of exposure to PFAS are substantial. Table 40: Estimated health-related costs of exposure to PFAS at different levels of exposure Exposure level Occupational (high) Elevated (medium) Background (low) Totals "Exposed" population and source of PFAS Workers at chemical production plants or manufacturing sites Communities near chemical plants, etc. with PFAS in drinking water Adults in general population (exposed via consumer products, background levels) Health endpoint linked to PFAS Kidney cancer All-cause mortality Low birth weight Infection Hypertension Nordic countries Population at risk n.a. Annual costs n.a. 621,000 EUR 2.1-2.4 billion 8,843 45,000 children 10.3 million 136 births of low birth weight 84,000 additional days of fever EUR 0.7-2.2 billion Nordic countries EUR 2.8-4.6 billion All EEA countries Population at risk 84,000- 273,000 Annual costs EUR 12.7-41.4 million 12.5 million 156,344 785,000 children 207.8 million EUR 41-49 billion 3,354 births of low birth weight 1.5 million additional days of fever EUR 10.7-35 billion All EEA countries EUR 52-84 billion The range of estimated annual health-related costs due to PFAS exposure is EUR 2.8-4.6 billion for the five Nordic countries and EUR 52-84 billion for all EEA countries. Some overlap occurs in the figures, because workers and affected communities are also exposed to background levels of PFAS. The actual costs are likely to be higher, since these calculations are for only a few of the health impacts linked to exposure to PFAS. 6.1.2 Non-health costs related to environmental contamination The second methodology compiled information on direct costs incurred by communities taking measures to reduce PFAS exposure through remediation of drinking water. Based on these direct costs, ranges of cost per persons affected or per case were developed. These unit costs then became the foundation for aggregating the costs of remediation for environmental exposure over and above action levels for PFAS concentrations in drinking water. As with the health-based estimates, the study assumes that 3% of the European population is exposed to drinking water with PFAS concentrations over regulatory action levels, such that the water treatment works serving them will require upgrading and maintenance over the next 20 years. Recognising the uncertainties that exist in the analysis and the available data, costs of remediation have been quantified using The cost of inaction linked to PFAS exposure 127 a scenario-based approach. For each scenario a number of parameters are specified, relating for example to the size of the affected population and the duration of maintenance works, and results generated accordingly. Table 41 shows the range of costs for the various categories of actions related to environmental remediation for the five Nordic countries. The overall range of estimated non-health costs is EUR 46 million - 11 billion over the next 20 years, just for the Nordic countries. The upper end of this range is dominated by soil remediation costs for which associated uncertainty must be considered high. Table 41: Detailed breakdown of ranges in quantified non-health costs for the Nordic countries. N people affected (3%) Screening and monitoring Health assessment Upgrade treatment works and maintenance Soil remediation Total Denmark 169,791 Finland 164,153 Iceland Norway 10,102 154,995 Sweden Nordic total 292,421 791,462 EUR 70,000- 8.3 million EUR 250,000- 22 million EUR 10,000- 900,000 EUR 170,000- 20 million EUR 480,000- 47 million EUR 280,000- 27 million EUR 270,000- 26 million EUR 20,000- 1.6 million EUR 260,000- 25 million EUR 490,000- 46 million EUR 7.4 million-274 million EUR 7.2 million- 265 million EUR 400,000- 1.6 million EUR 6.8 million- 250 million EUR 13 million- 472 million EUR 0-798 million EUR 2.2 million-2.1 billion EUR 100,000- 86 million EUR 1.6 million - 1.9 billion EUR 4.3 million - 4.5 billion EUR 8 million- 1.1 billion EUR 10 million- 2.4 billion EUR 1 million- 105 million EUR 9 million- 2.2 billion EUR 18 million- 5.1 billion EUR 46 million- 11 billion The cost estimates provided in the table are likely to be more robust at the aggregate, European level than at the national level, given the potential for significant variation between countries in sensitivity and use of PFAS, that could not be accounted for here. Parallel calculations for all 31 EEA Member Countries and Switzerland arrive at a range of non-health costs for environmental remediation totalling EUR 821 million to EUR 170 billion. Again, these cost estimates will be more robust at the aggregate, European level than at the national level. A review of the uncertainties concludes that the lower and upper bounds should be considered illustrative because of the limited information currently available, reflecting the level of variation in the unit cost estimates. However, based on the information from the literature review there is a firm basis for concluding that the lower bound estimates would be exceeded. A best estimate in the order of EUR 10-20 billion is certainly plausible. Significantly higher costs than that are likely if several cases similar to the contamination at Baden-Wurttemberg are identified, where costs of soil remediation have been estimated at up to EUR 3 billion. A number of other costs related to PFAS contamination are outside the scope of the quantification carried out in this report. These include loss of property value, reputational damage to a polluting company, costs of short-term measures such as provision of bottled water, ecological damage and the costs incurred by public authorities in responding to affected communities - including public outreach, surveys of contamination, and remedial measures. 128 The cost of inaction linked to PFAS exposure 6.1.3 Conclusions The work of estimating the health and environment-related costs to society related to PFAS exposure has relied on the development of assumption-based scenarios. This reflects the limited data available in the academic literature, government documents and press reports.Whilst the uncertainties of the analysis need to be acknowledged, it is also important to recognise that, for a number of issues, there is little or no uncertainty. For example, that the equivalent of hundreds of millions of EUR have already been spent on remediation of PFAS contamination, that PFAS use is widespread, and that PFAS will persist in the environment for an extremely long time. Other certainties include: 1. PFAS are ubiquitous in the environment, and almost all people have PFAS in their bodies today. Monitoring in both Sweden and the USA concludes that around 3% of the population are or have been exposed above proposed limit values, primarily through contamination of drinking water but also via other sources. 2. Many sources of PFAS exposure exist linked to specialist applictaions (e.g. AFFFs for firefighting at aairports and some industrial loctaions) and non-specialist uses (e.g. use in consumer goods such as ing, cosmetics and pizza boxes). 3. Non-fluorinated alternatives for many of these uses are already on the market, and therefore certain uses of PFAS can be reduced. 4. The costs for remediating some cases of contamination run to many millions of EUR. Total costs at the European level are expected to be in the hundreds of millions of EUR as a minimum. 5. A large and growing number of health effects have been linked to PFAS exposure and evidence is mounting that effects occur even at background level exposures. Current and proposed limit values for drinking water may be further reduced in recognition of growing information on, health and environmental risks. This would increase the costs of environmental remediation estimated here. As explained throughout the study, the calculations rest on a number of assumptions, though these have been checked against (e.g.) data on costs incurred to ensure that they are linked to real-world experience. As more information becomes available in the future, drawing on the framework provided here, calculations will become more precise. Moreover, these findings are conservative. The figures will only get larger, in that the numbers of PFAS on the market and the volumes produced keeps increasing. Further inaction will lead to more sources of contamination, more people exposed, and higher costs for remediation. The longer that PFAS contamination is left in the environment, the wider it will spread and the greater the quantity of soil or groundwater that will need to be decontaminated. The cost of inaction linked to PFAS exposure 129 6.2 Next steps and proposals for further studies As with other studies seeking to identify costs of inaction linked to chemicals exposure, our findings have been hampered by a lack of data concerning contamination and levels of human exposure. Overall calculations rest on a number of assumptions which may become more precise as more information is available in the future. Additional research to help gather this additional information could include: consideration of health endpoints due to exposures to groups of PFAS, since exposures are rarely limited to a single PFAS and since PFAS as a group share similar properties - most importantly, the property of extreme persistence in the environment; more information concerning the sites where production of PFAS and/or where manufacturing of manufacturing of products involving PFAS is, or have been occurring. National inventories of such sites are needed, including of sites where fluorinated AFFFs have been used. This would be of great help in estimating the numbers of affected populations, and the extent of contamination where remediation may be needed; systematic cataloguing of cost data where problems have been identified; inclusion of industries producing or using PFAS in the European Pollutant Release and Transfer Register, so that information on the location of releases to air and to water is available and so those releases can be tracked; national registries of products containing PFAS to help inform how PFAS are used and to contribute to better characterisation of the major sources of exposure from products; drinking water standards using group parameters for PFAS, so as to require better monitoring of drinking water. This will provide early warnings when elevated levels are found and enable more effective identification and timely containment of sources of contamination; better understanding of what happens to PFAS discharged from wastewater treatment plants and during incineration, including assessment of ecological risks from PFAS contamination; and more biomonitoring and epidemiological studies to characterize links between PFAS exposure and health endpoints, to enable better calculations of associated health costs. As this study highlights, the release of PFAS into the environment and constant exposure of humans is ongoing throughout the Nordic countries and Europe. Large-scale monitoring efforts such as those carried out in Sweden can help to clarify sources of contamination and provide more certainty concerning the scale of the socio-economic costs related to PFAS exposure. This will help to better inform policymakers, industry and consumers concerning the actions needed. 130 The cost of inaction linked to PFAS exposure References Adler A and Van der Voet J (2015). Occurrence and point source characterization of perfluoroalkyl acids in sewage sludge, Chemosphere 129: 62-73. Agency for Toxic Substances and Disease Registry (2018). Draft toxicological profile for perfluoroalkyls. Ahrens L (2011). Polyfluoroalkyl compounds in the aquatic environment: a review of their occurrence and fate. Journal of Environmental Monitoring, 13(1):20-31, as cited in California Department of Toxic Substances Control (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs. Discussion draft. Ahrens L et al. (2016). Screening of PFASs in groundwater and surface water. Uppsala Report No: 2016:2. Alder C (2018). 3M settles Minnesota groundwater lawsuit for $850 million. 21.02.2018. Accessed 09.09.2018. Alexandre G Paul et al. (2008). A First Global Production, Emission, And Environmental Inventory For Perfluorooctane Sulfonate, Environmental Science & Technology 2009 43 (2), 386-392. Almond D and Currie J (2011). Killing me softly: The fetal origins hypothesis. Journal of economic perspectives, 25(3), 153-72. Anderson H et al. (2016). Occurrence of select perfluoroalkyl substances at U.S. Air Force aqueous film-forming foam release sites other than fire-training areas: Field-validation of critical fate and transport properties, Chemosphere 150-678e685. Australia Government Department of Health (2018). Expert Health Panel for Per- and Poly-Fluoroalkyl Substances (PFAS). Avfall Sverige (2018). Hushllsavfall behandlad och insamlad mngd (In Swedish). Avfall Sverige (2018). PFAS on waste facilities ("PFAS p avfallsanlggningar"), Report no 2018:25 (in Swedish). Avinor (2018). PFOS I Focus, (In Norwegian) Accessed 05.09.2018. Bach C C et al. (2015). Perfluoroalkyl and polyfluoroalkyl substances and human fetal growth: a systematic review. Critical reviews in toxicology, 45(1), pp.53-67. Ballesteros V et al. (2017). Exposure to perfluoroalkyl substances and thyroid function in pregnant women and children: a systematic review of epidemiologic studies. Environ. Int.99, 15-28. doi: 10.1016/j.envint.2016.10.015. Banzhaf S et al. (2016). A review of contamination of surface-, ground-, and drinking water in Sweden by perfluoroalkyl and polyfluoroalkyl substances (PFASs). Ambio 46(3): 335-346. Barry V et al. (2013). Perfluorooctanoic acid (PFOA) exposures and incident cancers among adults living near a chemical plant. Environmental health perspectives, 121(11-12), p.1313. Becker M et al. (2008). Perfluoroactane surfactants in waste waters, the major source of river pollution, Chemosphere 72(1): 115-21. Doi: 10.1016/j.chemosphere.2008.01.009. Beekman M et al. (2016). Evaluation of substances used in the GenX technology by Chemours, Dordrecht, RIVM Letter. Report 2016-0174. Begley T H et al. (2008). Migration of fluorochemical-paper additives from food-contact paper into foods and food simulants. Food Additive and Contaminants: Part A 25(3):384-390. doi:10.1080/02652030701513784. Bharadwaj P et al. (2017). Birth Weight in the Long Run. Journal of Human Resources, 53(1), 189- 231. Black and Veatch (2018) Alternatives Evaluation Report. Emerging contaminants treatments strategies study. The cost of inaction linked to PFAS exposure 131 Black S et al. (2007). From the cradle to the labor market? The effect of birth weight on adult outcomes. The Quarterly Journal of Economics, 122(1), 409-439. Black and Veatch (Report prepared for Cape Fear Public Utility Authority) (2018). Alternatives Evaluation Report, Emerging Contaminants Treatment Strategies Study. Project no: 196369. Blaine A C et al. (2013). Uptake of Perfluoroalkyl Acids into Edible Crops via Land Applied Biosolids: Field and Greenhouse Studies. Environmental Science & Technology, 47(24), pp. 14062-14069. DOI: 10.1021/es403094q. Blum A et al. (2015). The Madrid Statement on Poly- and Perfluoroalkyl Substances (PFASs). Environmental health perspectives. Vol. 123, no. 5. Boele B Schiphol Logistics Park blijkt vervuild, Haarlems Dagblad, January 3 2018. Accessed 02.09.2018. Bonefeld-Jrgensen E C et al. (2011). Perfluorinated compounds are related to breast cancer risk in Greenlandic Inuit: a case control study. Environmental Health, 10(1), p.88. Bonefeld-Jrgensen E C et al. (2014). Breast cancer risk after exposure to perfluorinated compounds in Danish women: a case-control study nested in the Danish National Birth Cohort. Cancer Causes & Control, 25(11), pp.1439-1448. Brandveilig (2011). Zorgen om verontreiniging door incident met sprinklerinstallatie. Accessed 08.09.2018. Breast Cancer Fund (2015). Anti-aging secrets exposed: chemical linked to breast cancer found in skin care. Accessed 15.09.2018. Brede E et al. (2010). Two-year follow-up biomonitoring pilot study of residents' and controls' PFC plasma levels after PFOA reduction in public water system in Arnsberg. International journal of hygiene and environmental health, 213(3), pp.217-223. doi.org/10.1016/j.ijheh.2010.03.007. Busch J (2010). Polyfluoroalkyl compounds in landfill leachates. Environmental Pollution, Vol 158:5, 1467-1386. California Department of Toxic Substances Control (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs (Discussion draft). CEFIC (2017). Chemical Industry Profile. Accessed 18.09.2018. Center for Disease Control and Prevention (2018) Fourth National Report on Human Exposure to Environmental Chemicals. Updated Tables Volume 1. Center for Environmental Health (2018). Avoiding Hidden Hazards. Chemical Watch (2018). Cosmetics giant L'Oral to eliminate PFASs in products. Accessed 02.09.2018. Chemours (2018). The Chemours Company Takes Significant Action to Minimize Emissions. Accessed 08.09.2018. Clara M et al. (2008). Emissions of perfluorinated alkylated substances (PFAS) from point sourcesidentification of relevant branches. Water Science and Technology 58(1):59-66. doi: 10.2166/wst.2008.641. Climate and Pollution Agency (2012). Inventory of PFOS and PFOS-related substances in firefighting foams in Norway. Commission Regulation (EU) No 1907/2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC. ("REACH Regulation"), O.J. 396, 30.12.2006. Concawe (2016). Environmental fate and effects of poly- and perfluoroalkyl substances (PFAS). Report no: 8/16. Consonni et al. (2013). Cancer Risk among TFT synthesis and polymerization workers, American Journal of Epidemiology, DOI: 10.1093/aje/kws588. Copp, T, DoD: At least 126 bases report water contaminants linked to cancer, birth defects, Military Times, 26 April 2018. Accessed 01.11.2018. 132 The cost of inaction linked to PFAS exposure Costa G et al. (2009). Thirty years of medical surveillance in perfluorooctanoic acid production workers. Journal of Occupational and Environmental Medicine 51:364-372. Council Directive of 12 December 1991 on hazardous waste (91/689/EEC). Currie J and Hyson R (1999). Is the impact of health shocks cushioned by socioeconomic status? The case of low birthweight. American Economic Review, 89(2), 245-250. Danish Environmental Protection Agency (2011). Substitutions of PFOS for use in non-decorative hard chrome plating. Danish Environmental Protection Agency (2014). Screeningsundersgelse af udvalgte PFASforbindelser som jord- og grundvandsforurening I forbindelse med punktkilder (In Danish). Danish Environmental Protection Agency (2015). Perfluorerede alkylsyreforbindelser (PFAS-forbindelser) incl. PFOA, PFOS og PFOSA. Danish Environmental Protection Agency (2016). Affaldsstatistikken 2016. (In Danish) Danish Environmental Protection Agency (2016). Kortlgning af brancher der anvender PFAS. Dalsager L et al. (2016). Association between prenatal exposure to perfluorinated compounds and symptoms of infections at age 1-4years among 359 children in the Odense Child Cohort. Environment international, 96, pp.58-64. Danish Environmental Protection Agency (2016). Survey and risk assessment of chemical substances in rugs for children. Danish Environmental Protection Agency (Not yet published). Risk Assessment of fluorinated substances in cosmetic products. Darrow L A et al. (2014). PFOA and PFOS serum levels and miscarriage risk. Epidemiology, 25(4), pp.505-512. Darrow L A et al. (2016). Modeled perfluorooctanoic acid (PFOA) exposure and liver function in a mid-Ohio valley community. Environmental health perspectives, 124(8), p.1227. Darwin R (for Fire Fighting Foam Coalition) (2011). Estimated Inventory of PFOS-based Aqueous Film Forming Foam (AFFF). Arlington, Virginia. DeMasi M, Hoosick Falls water filtration still weeks away, Albany Business Review, Jan 29, 2016. Accessed 05.08.2018. Dhingra R et al. (2017). A study of reverse causation: Examining the associations of perfluorooctanoic acid serum levels with two outcomes. Environmental health perspectives, 125(3), p.416. ECHA (2014). Germany and Norway propose a restriction on Perfluorooctanoic acid (PFOA), its salts and PFOA related substances. ECHA (2016). Valuing selected health impacts of chemicals: Summary of the results and a critical review of the ECHA study. ECHA (2017a). Inclusion of substances of very high concern in the Candidate List for eventual inclusion in Annex XIV. ECHA (2017b). Public consultation. Germany, in collaboration with Sweden, proposed a restriction on C9-C14 perfluorocarboxylic acids (PFCAS), their salts and related substances (precursors). Editorial, We must take control of our water quality, Starnews Online, 25.06. 2018. Accessed 11. 09. 2018. EFSA Panel on Contaminants in the Food Chain, Knutsen HK et al., (2018). Scientific opinion on the risk to human health related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. EFSA Journal 16(12):5194, doi:10.2903/j.efsa.2018.5194 Ellison G (2018). Belmont woman's blood is 750 times national PFAS average, Mlive, 11.01.2018 Accessed 15.10.2018. Environment Protection Agency (2016). Fact Sheet PFOA& PFOS Drinking Water Health Advisories. Environment Protection Agency (2016). Health effects support document for perfluorooctanoic acid (PFOA). Document no: EPA822R16003. Environment Protection Agency (2016). Health effects support document for perfluorooctane sulphate (PFOS). Document No: EPA822R16002. The cost of inaction linked to PFAS exposure 133 Environment Protection Agency (2018). Advancing Sustainable Materials Management: 2015 Fact Sheet. Environment Protection Agency (2018). Fact Sheet: Draft Toxicity Assessments Environment Protection Agency (2018). Historic EPA Summit Provides Active Engagement and Actions to Address PFAS. Environmental Protection Agency of Denmark, Risk Assessment of fluorinated substances in cosmetic products (Not yet published). Ericson I et al. (2007). Perfluorinated chemicals in blood of residents in Catalonia (Spain) in relation to age and gender: a pilot study. Environ. Int. 33, 616-623. DOI: 10.1016/j.envint.2007.01.003. Eriksson U et al. (Report for Swedish Environmental Protection Agency) (2015). Screening of PFASs in sludge and water from waste water treatment plants. Umea Sweden. Eschauzier C et al. (2012). Polyfluorinated Chemicals in European Surface Waters, Ground and Drinking Waters. In: Knepper T., Lange F. (eds) Polyfluorinated Chemicals and Transformation Products. The Handbook of Environmental Chemistry, vol 17. Springer, Berlin, Heidelberg. European Heart Network (2017). European Cardiovascular Disease Statistics 2017. Eurostat Transport Statistics [AIRP_TYP]. Accessed: 14.06.2018. Eurostat, annual enterprise statistics by size class for special aggregates of activities (NACE Rev. 2) [sbs_sc_sca_r2]. Eurostat Database, Wastewater treatment plants by treatment level [env_ww_plt]. EWG Cosmetics Database, Is Teflon in your Cosmetics? Published March 14, 2018. Federal Environmental Agency (2017). Fortschreibung der vorlufigen Bewertung von per- und polyfluorierten Chemikalien (PFC) im Trinkwasser. Bundesgesundheitsbl. 2017 60:350-352. (In German). Field J et al. (Report for Environmental Security Technology Certification Program-ESTCP) (2017). FAQs Regarding PFASs Associated with AFFF Use at U.S. Military Sites. Fischer S (2017). Known uses of PFASs (presentation at Nordic workshop on joint strategies for PFAS). Fischer S et al. (2016). Poly- and perfluoroalkyl substances on the market and in the Swedish environment. Norman Bulletin Issue 5. Foundation for Water Research (2008). Survey of the prevalence of perfluorooctane sulphonate (pfos), perfluorooctanoic acid (pfoa) and related compounds in drinking water and their sources. Franko et al. (2012). Dermal Penetration potential of Perfluorootanoic Acid (PFOA) in Human and Mouse Skin. Journal of Toxicology and Environmental Health, Part A. 75:50-62. Frisbee SJ et al. (2009). The C8 health project: design, methods, and participants. Environmental health perspectives, 117(12), 1873-82. doi: 10.1289/ehp.0800379. Fuchs Dr. Jacques G (For FiBL) (2008). Compost and digestate: sustainability, benefits, impacts for the environment and for plant production. Fuertes I et al. (2017), Perfluorinated alkyl substances (PFASs) in northern Spain municipal solid waste landfill leachates, Chemosphere 168: 399-407. Fujii et al. (2013). Occurrence of perfluorinated carboxylic acids (PFCAs) in personal care products and compounding agents. Chemosphere 93 (2013) 538-544. Gabbert S (2018). Economic assessments and valuations of environmental and health impacts caused by Perfluorooctanoic acid (PFOA) and its salts. OECD Environment Working Paper No 128. Gardner P, Astronomical' PFAS level sets new Michigan contamination milestone, Mlive , 24.09.2018. Accessed 05.11.2018. German Environment Agency (2011). Grenzwerte, Leitwerte, Orientierungswerte, Manahmenwerte Aktuelle Definitionen und Hchstwerte (In German). German Environment Agency (2017). Use of PFOS in chromium plating - Characterisation of closed-loop systems, use of alternative substances. Report No. (UBA-FB) 002369/ENG. 134 The cost of inaction linked to PFAS exposure German Environmental Agency (2017). Fortschreibung der vorlufigen Bewertung von per- und polyfluorierten Chemikalien (PFC) im Trinkwasser (In German). Bundesgesundheitsbl 2017 60:350-352. Gesundheitsamt Rastatt (No date) Informationen zum Stand der PFC-Blutkontrolluntersuchung im Landkreis Rastatt. Gille J (2018). Swedish Armed Forces' Sustainability Section, personal communication, September 2018. Gomis M I (2018). Comparing the toxic potency in vivo of long-chain perfluoroalkyl acids and fluorinated alternatives. Environment international, 113. Grandjean P et al. (2012). Serum vaccine antibody concentrations in children exposed to perfluorinated compounds. Jama, 307(4), pp.391-397. Grandjean P (2018). Delayed discovery, dissemination, and decisions on intervention in environmental health: a case study on immunotoxicity of perfluorinated alkylate substances. Environmental Health (2018) 17:62. doi.org/10.1186/s12940-018-0405-y. Grandjean P and Clapp R (2015). Perfluorinated alkyl substances: emerging insights into health risks. New solutions: a journal of environmental and occupational health policy, 25(2), pp.147-163. Grandjean P et al. (2014). Changing interpretation of human health risks from perfluorinated compounds. Public health reports, vol. 129: (6). Grice M et al. (2007). Self-reported medical conditions in perfluorooctanesulfonyl fluoride manufacturing workers. Journal of Occupational and Environmental Medicine 49:722-729. DOI: 10.1097/JOM.0b013e3180582043. Hamid H and Li L (2016). Role of wastewater treatment plant (WWTP) in environmental cycling of poly- and perfluoroalkyl (PFAS) compounds. Ecocycles 2(2), as cited in California Department of Toxic Substances Control (2018). Product - Chemical Profile for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs) in Carpets and Rugs. Discussion draft. Hamid H et al. (2018). Review of the fate and transformation of per- and polyfluoroalkyl substances (PFASs) in landfills. Environmental Pollution, 235: 74-84. https://doi.org/10.1016/j.envpol.2017.12.030. Hansson et al. (2016). Sammanstllning av befintlig kunskap om froreningskllor till PFAS-mnen i svensk milj Report Number C 182. Haug L S et al. (2009). Time Trends and the Influence of Age and Gender on Serum Concentrations of Perfluorinated Compounds in Archived Human Samples. Environ. Sci. Technol. 43:6, 2131- 2136. DOI: 10.1021/es802827u. Heggelund A (2018). Norwegian EPA. Personal communication. Henricsson C (Master Thesis) (2017). The Presence of PFAS in Cosmetic Products (Frekomst av PFAS i kosmetiska produkter). University of Lund. Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). Holzer J et al. (2008). Biomonitoring of perfluorinated compounds in children and adults exposed to perfluorooctanoate (PFOA) contaminated drinking water. Environ. Health Perspect. 116, 651- 657. DOI: [10.1289/ehp.11064]. Hlzer J et al. (2009). One-year follow-up of perfluorinated compounds in plasma of German residents from Arnsberg formerly exposed to PFOA-contaminated drinking water, International journal of hygiene and environmental health. 212(5), pp.499-504. Hopkins et al. (2018). Recently Detected Drinking Water Contaminants: GenX and Other Per- and Polyfluoroalkyl Ether Acids. American Water Works Association Journal 110:7 p.13-28. Houtz E F et al. (2016). Poly- and perfluoroalkyl substances in wastewater: Significance of unknown precursors, manufacturing shifts, and likely AFFF impacts. Water Research 95: 142-9. Doi:10.1016/j.watres.2016.02.055. Huber S et al. (2009). Emissions from incineration of fluoropolymer materials, Report number: OR112/2009. IARC (2018). Perfluorooctanoic acid (PFOA). The cost of inaction linked to PFAS exposure 135 Ingelido A M et al. (2018). Biomonitoring of perfluorinated compounds in adults exposed to contaminated drinking water in the Veneto Region, Italy. Environment international, 110, pp.149-159. International Association of Fire and Rescue Services (2017). World Fire Statistics. International Car Wash Association. Industry Information. Interstate Technology Regulatory Council (2017). History and use of per- and polyfluoroalkyl substances. Interstate Technology Regulatory Council (no date). Environmental Fate and Transport for Perand Polyfluoroalkly Substances. IPEN (2018). Fluorine-free firefighting foams (3f) viable alternatives to fluorinated aqueous filmforming foams (AFFF). ISS (National Institute of Health). Dept of Environmental Health (no date) Perfluorooctanoic acid (PFOA) pollution of groundwater: the case study of Veneto, Italy. Jacobsen J B and Hanley N (2009). Are There Income Effects on Global Willingness to Pay for Biodiversity Conservation? Environmental and Resource Economics, Volume 43, Issue 2, pp 137-160. Jakobsson K et al. (2014). Exponering fr perfluorerade mnen (PFAS) i dricksvatten i Ronneby kommun (In Swedish). Rapport 8:2014. Jamberg U and Holmstrom, K (2007). Perfluoroalkylated acids and related compounds (PFAS) in the Swedish environment, Stockholm University. Jensen T K (2015). Association between perfluorinated compound exposure and miscarriage in Danish pregnant women. PloS one, 10(4), p.e0123496. Jersey Airport Website dedicated to the PFOS, available here. Kannan K et al. (2004). Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ. Sci. Technol. 38, 4489-4495. Krrman A et al. (2004). Levels of perfluoroalkylated compounds in whole blood from Sweden. Organohalogen Compd. 66, 4008-4012. Krrman A et al. (2011). Environmental levels and distribution of structural isomers of perfluoroalkyl acids after aqueous fire-fighting foam (AFFF) contamination. Environ. Chem., 8, 372-380. Kary T (2018). 3M Settles Minnesota Lawsuit for $850 million. Bloomberg, 20.02.2018. Accessed 08.13.2018. Kato K et al. (2011). Trends in exposure to polyfluoroalkyl chemicals in the US population: 1999- 2008. Environmental science & technology, 45(19), pp.8037-8045. Kennebunk, Kennebunkort and Wells Water District (2017). Per and Polyfluoroalkyl Substances (PFASs) Usage (Draft for informational purposes). Kissa, E (2001). Fluorinated Surfactants and Repellents. Surfactant Science Series. 97. Marcel Dekker, New York. Kjeldsen L S and Bonefeld-Jrgensen EC (2013). Perfluorinated compounds affect the function of sex hormone receptors. Environmental Science and Pollution Research, 20(11), pp.8031-8044. Klatt P and Frey A (2016). Woher kam das Zeug blo?, Frankfurter Allgemeine, 4 September 2016. Accessed September 2018. Kotterman MJJ and Kwadijk C.J.A.F (2009). PFOS onderzoek in waterbodem en vis. Report No: C064/09 (in Dutch). Kotthoff M et al. (2015). Perfluoroalkyl and polyfluoroalkyl substances in consumer products. Environmental Science and Pollution Research International. 22(19): 14546-14559. Krista A. Barzen-Hanson et al. (2017). Discovery of 40 classes of per- and polyfluoroalkyl substances in historical aqueous film-forming foams (AFFFs) and AFFF-impacted groundwater. Environmental Science and Technology 2017, 51 (4), pp 2047-2057 DOI: 10.1021/acs.est.6b05843. KRRC (2016). Air Force Signs Contract for Bottled Water Distribution. KRRC (2016). Health & Water Officials Try to Reassure Residents in Areas of PFC Contamination. Lang JR et al. (2016). Release of Per- and Polyfluoroalkyl Substances (PFASs) from Carpet and Clothing in Model Anaerobic Landfill Reactors. Environmental Science & Technology, 50, 5024- 5032. 136 The cost of inaction linked to PFAS exposure Lang JR et al. (2017). National Estimate of Per- and Polyfluoroalkyl Substance (PFAS) Release to U.S. Municipal Landfill Leachate. Environmental Science and Technology, 21;51(4):2197-2205. doi: 10.1021/acs.est.6b05005. Laurel A et al. (2017). Fluorinated Compounds in U.S. Fast Food Packaging. Environmental Science & Technology Letters. doi: 10.1021/acs.estlett.6b00435. Leal J (2016). Economic burden of bladder cancer across the European Union. European urology, 69(3), pp.438-447. Lee L and Trim H (2018). Summary Sheet Evaluating Perfluoroalkyl Acids in Composts with Compostable Food Serviceware Products in their Feedstocks. Lerner S (2016). New Teflon Toxin Causes Cancer in Lab Animals, The Intercept, 03.06.2016. Accessed 18.09.2018. Lerner S (2018), 3M knew about the dangers of PFOA and PFOS decades ago, internal documents show, The Intercept, 31.07.2018. Accessed 18.09.2018. Li Ying et al. (2017). Technical Report: Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. Loos et al. (2009). EU wide survey of polar organic persistent pollutants in European river waters. Environmental Pollution, 157: 561-568, as cited in the Concawe report (2016). Environmental fate and effects of poly and perfluoroalkyl substances (PFAS). Loos et al. (2013). as cited in Concawe report (2016). Environmental fate and effects of poly and perfluoroalkyl substances (PFAS). Loos R et al. (2010). Pan-European survey on the occurrence of selected polar organic persistent pollutants in ground water. Water Research 44: 4115-4126, as cited in the Concawe report (2016). Environmental fate and effects of poly and perfluoroalkyl substances (PFAS). LUBW (2017). Grundwasserberwachungsprogramm Ergebnisse der Beprobung 2017 (in German). Luengo-Fernandez R (2013). Economic burden of cancer across the European Union: a populationbased cost analysis. The lancet oncology, 14(12), pp.1165-1174. Marcotty J (2017). State alleges 3M chemicals caused cancer and infertility, alleges $5 billion in damage. Star Tribune, 21.11.2017. Accessed 10.10.2018. Marcotty J (2018) 3M settles groundwater lawsuit for $850 million, Star Tribune, February 20 2018. Accessed 15.10.2018. Marten Law (website). The True Cost of Scotchgard: 3M to Pay Minnesota $850 Million in Perfluorochemical Settlement. Accessed 10.10.2018. Mastrantonio M et al. (2017). Drinking water contamination from perfluoroalkyl substances (PFAS): an ecological mortality study in the Veneto Region, Italy. The European Journal of Public Health. Feb 1;28(1):180-185. doi: 10.1093/eurpub/ckx066. Mata-Cases M et al. (2016). Direct medical costs attributable to type 2 diabetes mellitus: a populationbased study in Catalonia, Spain. The European Journal of Health Economics, 17(8), pp.1001-1010. Min, J. Y. et al. (2012). Perfluorooctanoic acid exposure is associated with elevated homocysteine and hypertension in US adults. Occup Environ Med, 69(9):658-62. doi: 10.1136/oemed-2011-100288. Minister of Defence Hultqvist P. (2017) Answer to question on removal of PFAS. Dnr F2017/01275/MFI. Ministerium fr Umwelt, Klima und Energiewirtschaft Baden-Wrttemberg (2018). Neuer Erlass schreibt Beurteilungsgrundlage fr mgliche PFC-Belastungen von Grund- und Sickerwasser fort Ministerium fr Umwelt, Klima und Energiewirtschaft Baden-Wrttemberg (No date). PFCBelastung in Baden-Baden und im Landkreis Rastatt. Ministry of Environment and Food of Denmark (2015). Perfluorerede alkylsyreforbindelser (PFASforbindel Ser) incl. PFOA, PFOS og PFOSA. Ministry of Infrastructure and Environment of Netherlands and Public Waste Agency of Flanders (2016). Inventory of awareness, approaches and policy: Insight in emerging contaminants in Europe. Minnesota Department of Health Perfluoroalkyl Substances (PFAS) accessed: 09.10. 2018. Miteni Company Website: http://www.miteni.com/index.htm The cost of inaction linked to PFAS exposure 137 Mordock J, C8 suspected in birth defects: One woman's story, Delaware Online, April 2 2016. Accessed 12.10.2018. Nair A S, DuPont settles lawsuits over leak of chemical used to make Teflon. Reuters Business News, February 13 2017. Accessed 10.09.2018. National Cancer Institute [NCI] Surveillance, Epidemiology, and End Results data for 2005-2011: https://seer.cancer.gov/././ National Research Council (1997). Innovations in Ground Water and Soil Cleanup: From Concept to Commercialization. Chapter 6. National Academy Press. Navrud S and Ready R (eds.) (2007). Environmental Value Transfer: Issues and Methods. Springer. Nelson J et al. (2010). Exposure to polyfluoroalkyl chemicals and cholesterol, body weight, and insulin resistance in the general US population. Environmental health perspectives, 118(2), 197. DOI: 10.1289/ehp.0901165. Neltner T and Maffini M (2018). Paper mills as a significant source of PFAS contamination, but who's watching? EDF, 21.05.2018. Accessed 10.11.2018. New Jersey Register, adopted Amendments: N.J.A.C. 7:9E-2.1; 7:10-5.2, 5.3, and 12.30; and 7:186.4. No Author. Another Chapter: Chemours releases GenX and PFOAs into waters globally, Encore, August 8 2018. Accessed 10.10.2018. No Author. Bodemvervuiling bij Schiphol snel aanpakken, De Volkskrant, 12 October 2011, Accessed 05.09.2018. No author, La procura di Vicenza indaga sulle sostanze cancerogene nell'acqua, Le Iene, 21.05. 2018, Accessed October 2018 (in Italian). Norden (2013). Per- and polyfluorinated substances in the Nordic Countries: Use, occurrence and toxicology. TemaNord 2013:542. Nordic Council of Ministers (2014). The Cost of Inaction - a socioeconomic analysis of costs linked to effects of endocrine disrupting substances on male reproductive health. TemaNord 2014:557. Nordic Council of Ministers (2017). Analysis of PFAS and TOF in products. Nordic Ecolabelling (2018). About Nordic SAWN Ecolabelled Cosmetic Products. Version 3.3. Norwegian Environment Agency (2016). PFAS-forurensning i grunnen Oppsummering fra workshop 26. Rapport M-622. OECD (2008). OECD Environmental Outlook to 2030, OECD Publishing, Paris, https://doi.org/10.1787/9789264040519-en. OECD (2012). Mortality risk valuation in environment, health and transport policies. OECD (2016). Health status statistics. OECD (2018). Toward a new comprehensive global database of per and polyfluoroalkyl substances (pfass): summary report on updating the OECD 2007 list of per and polyfluoroalkyl substances (PFASs). Oliaei F et al. (2013). PFOS and PFC releases and associated pollution from a production plant in Minnesota (USA). Environmental Science Pollution Research. 20: 1977-1992. Olsen G W (2001). A longitudinal analysis of serum perfluorooctanesulfonate (PFOS) and perfluorooctanonate (PFOA) levels in relation to lipid and hepatic clinical chemistry test results from male employee participants of the 1994/95, 1997 and 2000 fluorochemical medical surveillance program. Final report Epidemiology, 220-3W-05, Medical department, 3M Company Olsen G W et al. (2003). Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations. Journal of Occupational and Environmental Medicine, 45(3), pp.260-270. More information about this controversy can be found in this news article: https://theintercept.com/2018/07/31/3m-pfas-minnesota-pfoa-pfos/ Onyshko, J and Hewlett R (Anthesis Consulting Group) (2018). Toxics in carpet in the European Union. Oosterhuis F and Brouwer R (2015). Benchmark Development for the proportionality assessment of PBT and vPvB substances. IVM Institute for Environmental Studies. Place B and Field J (2012). Identification of Novel Fluorochemicals in Aqueous Film-Forming Foams Used by the US Military. Environ. Sci. Technol. 46 (13). pp 7120-7127 DOI: 10.1021/es301465n. 138 The cost of inaction linked to PFAS exposure Press release (No date) Fluorotelomers Market to Reach USD 539.3 Million Worldwide by 2020, Digital Journal. Accessed 10.11.2018. Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on the quality of water intended for human consumption COM/2017/0753 final - 2017/0332 (COD). RAC/SEAC (2016) Background Document to the Opinion on the Annex XV dossier proposing restrictions on Octamethylcyclotetrasiloxane (D4) and Decamethylcyclopentasiloxane (D5). Committee for Risk Assessment and Committee for Socio-Economic Analysis, European Chemicals Agency, Helsinki. Rappazzo K et al. (2017). Exposure to perfluorinated alkyl substances and health outcomes in children: a systematic review of the epidemiologic literature. International journal of environmental research and public health, 14(7). p.691. Jun 27;14(7). pii: E691. doi: 10.3390/ijerph14070691. Regierungsprsidium Karlsruhe (2017). Antworten auf hufig gestellte Fragen zur PFC-Belastung im Landkreis Rastatt und den Stadtkreisen Baden-Baden und Mannheim Stand August 2017. Regierungsprsidium Karlsruhe (No date). berblick zur PFC-Problematik in Mittel- und Nordbaden. Accessed September 2018. Regione di Veneto (2018). Piano di Sorveglianza Sanitaria Sulla Popolazione Esposte alle Sostanze Perfluoroalchiliche (in Italian), Rapport no 04, p.10. Reisch M S (2018). Merged lawsuit filed against DuPont and Chemours in North Carolina, Chemical and Engineering News, February 5 2018, Accessed 08.09.2018. Reuning A, Landschaft mit Gift, Deutschland Funk, 23.04.2017. Accessed 10.11.2018. Ritscher A et al. (2018). Zrich Statement on Future Actions on Per- and Polyfluoroalkyl Substances (PFASs), Environmental Health Perspectives, vol 126, no 8. RIVM (2009). Estimation of emissions and exposures to PFOS used in industry. Report 601780002/2009. RIVM (2016). GenX. Accessed September 2018. RIVM (2016). Risicoschatting emissie PFOA voor omwonenden (In Dutch). RIVM Briefrapport 20160049. RIVM (2018). Risicobeoordeling van GenX en PFOA in moestuingewassen in Dordrecht, Papendrecht en Sliedrecht (In Dutch). Robins N S (British Geological Survey)(2000). The water resources of Jersey: an overview. Ronneby Municipality (No date). Frgor och svar om PFAS. Last updated 17.07.2018. Accessed 08.08.2018. RPA and BRE Environment (2004). Perfluorooctane Sulphonate: Risk reduction strategy and analysis of advantages and drawbacks. Report no: J454/PFOS RRS. Sahlin S (2017). PFAS in the Baltic Sea Region. Saikat S et al. (2013). The impact of PFOS on health in the general population: a review. Environmental Science: Processes & Impacts, 15(2), pp.329-335. Scher D et al. (2018). Occurrence of perfluoroalkyl substances (PFAS) in garden produce at homes with a history of PFAS-contaminated drinking water. Chemosphere 196: 548-555. Scheringer M et al. (2014). Helsingr Statement on poly- and perfluorinated alkyl substances (PFASs)', Chemosphere, vol. 114. Schiphol Airport (2017). Annual Report (In Dutch). Schultz, M et al. (2003). Fluorinated alkyl surfactants. Environmental Engineering Science 20(5) 487- 501. Schyberg, I. (2018). Ronneby Milj och Teknik AB, personal communication. Scott A (2016). Dutch Chemical Plant Under Investigation, Chemical and Engineering News, April 18 2016. Accessed 10.11.2018. Sedlak et al. (2017). Per- and polyfluoroalkyl substances (PFASs) in San Francisco Bay wildlife: Temporal trends, exposure pathways, and notable presence of precursor compounds. Chemosphere 185:1217-1226. Doi:10.1016/j.chemosphere.2017.04.096. The cost of inaction linked to PFAS exposure 139 Shankar A et al. (2012). Perfluorooctanoic acid and cardiovascular disease in US adults. Archives of Internal Medicine, 172(18), pp.1397-1403. DOI: 10.1001/archinternmed.2012.3393. Skutlarek D et al. (2006). Perfluorinated surfactants in surface and drinking waters. Environmental science and pollution research international, 13(5), p.299. Shlmann, R. (2018). Landratsamt Rastatt, personal communication. Sportvisserij Netherlands (News) Rijnland hervat waterafvoer Schiphol terrain, 21 July 2008, Accessed: August 2018. (In Dutch). Stadtwerke Rastatt (2018). Rodriguez, L. Personal communication, December 2018. Stadtwerke Rastatt (2018). So schtzen wir unser Trinkwasser in Rastatt (website in German) Stale Navrud (2016). Possibilities and challenges in transfer and generalization of monetary estimates for environmental and health benefits of regulating chemicals. (Presentation at OECD Workshop on socioeconomic impact assessment of chemicals management). State of Michigan, Website dedicated to PFAS contamination. States of Jersey. States Greffe Logs, Jersey Airport: fireground remediation - Deed of settlement, 19 October 2004. Accessed 08.09.2018. Statistisk sentralbyr (2018). 426 kilo avfall per innbyggjar (In Norwegian) Steenland K (2013). Ulcerative colitis and perfluorooctanoic acid (PFOA) in a highly exposed population of community residents and workers in the mid-Ohio valley. Environmental health perspectives, 121(8). Steenland K and Woskie S (2012). Cohort mortality study of workers exposed to perfluorooctanoic acid. American journal of epidemiology. 176(10) pp.909-917. Steenland K et al. (2009). Association of perfluorooctanoic acid and perfluorooctane sulfonate with serum lipids among adults living near a chemical plant. American journal of epidemiology, 170(10), pp.1268-1278. Steenland K et al. (2010). Association of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) with uric acid among adults with elevated community exposure to PFOA. Environmental health perspectives, 118(2). p.229. Stein C R et al. (2009). Serum levels of perfluorooctanoic acid and perfluorooctane sulfonate and pregnancy outcome. American journal of epidemiology, 170(7), pp.837-846. Stockholm Convention (2018). Fourteenth meeting of the Persistent Organic Pollutants Review Committee (POPRC.14) . Accessed December 2018. Stockholm Convention (website). POPRC Recommendations for listing Chemicals. Accessed December 2018. Sullivan M, Car wash cites for PFAS pollution, Sea Coast Online, June 4 2018. Accessed 19 August 2018. Sunding D L (2017). Damage to Minnesota's natural resources resulting from 3M's disposal of PFCs in Washington County. (Report prepared for the State of Minnesota in the matter of the State of Minnesota v. 3M Company.) Swanson M, People, not politics, must be PFAS priority, The Herald, May 2018. Accessed 12.11.2018. Swedavia (2017) Environmental Debt report 2017 ("Miljskuldsrapport") (In Swedish) Swedavia Airports, (2018). PFAS p Swedavias flygplatser. Swedish Chemicals Agency (2014). Chemical Analysis of Selected Fire-fighting Foams on the Swedish Market Swedish Chemicals Agency (2015). Occurrence and use of highly fluorinated substances and alternatives: Report from a government assignment. Report 7/15. Swedish Chemicals Agency (2015). Survey of Fire Fighting Foam. Report 5/15. Swedish Chemicals Agency (2018). (News) The Swedish Chemicals Agency is introducing a requirement to report PFASs to the Products Register. Accessed September 2018. Swedish Environmental Protection Agency (2016). Hgfluorerade mnen (PFAS) och bekmpningsmedel en sammantagen bild av frekomsten i miljn. 140 The cost of inaction linked to PFAS exposure Swedish Environmental Protection Agency (2018). Frdjupad miljvervakning av hgfluorerade miljgifter (s.k. PFAS) och av vxtskyddsmedel i vatten NV-08978-16 (In Swedish). Swedish Environmental Protection Agency (2015). Risks and Effects of the dispersion of PFAS on Aquatic, Terrestrial and Human populations in the vicinity of International Airports: Final Report of the RE-PATH project. Report number: B 2232. Swedish National Food Agency (2017). Riskhantering - PFAS i dricksvatten och fisk (In Swedish) Swedish Water & Wastewater Association (2016). Produktion av dricksvatten. Swedish Environmental Protection Agency (2016). Highly fluorinated substances (PFAS) and pesticide, an overview of the presence in the environment (report in Swedish with English Summary). Tang, C Y et al. (2007). Effect of Flux (Transmembrane Pressure) and Membrane Properties on Fouling and Rejection of Reverse Osmosis and Nanofiltration Membranes Treating Perfluorooctane Sulfonate Containing Wastewater. Environmental Science and Technology 41: 2008-2014. Taylor P S et al. (2014). Investigation of waste incineration of fluor4otelomer-based polymers as a potential source of PFOA in the environment. Chemosphere 110: 17-22. The Commission Regulation (EU) 2017/1000 of 13 June 2017 amending Annex XVII to Regulation (EC) No 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) as regards perfluorooctanoic acid (PFOA), its salts and PFOA-related substances. The European Commission-DG Enterprise and Industry (2010). Analysis of the risks arising from the industrial use of Perfluorooctanoic acid (PROA) and Ammonium Perfluorooctanoate (APFO) and from their use in consumer articles. The Government of Netherlands (2018). Beantwoording Kamervragen over berichten over met PFOS en PFOA verontreinigde grond bij Schiphol. (In Dutch). The Groundwater Association (2017). PFAS Top 10 Facts. The National Academies of Science Engineering and Medicine (2017). Use and Potential Impacts of AFFF Containing PFASs at Airports. Tittlemier SA et al. (2007). Dietary exposure of Canadians to perfluorinated carboxylates and perfluorooctane sulfonate via consumption of meat, fish, fast foods, and food items prepared in their packaging. Journal of Agricultural and Food Chemistry. 55(8):3203-3210; doi:10.1021/jf0634045. Trasande L et al. (2015). Estimating burden and disease costs of exposure to endocrine-disrupting chemicals in the European Union. J Clin Endocrinol Metab. 2015; 100: 1245-1255. Trier X et al. (2011). Polyfluorinated surfactants (PFS) in paper and board coatings for food packaging. Environ Sci Pollut Res Int 18(7):1108-1120. doi:10.1007/s11356-010-0439-3. Trudel D et al. (2008). Estimating Consumer Exposure to PFOS and PFOA. Risk Analysis: An International Journal 28.2: 251-269. Turchetti, G et al. (2017). The social cost of chronic kidney disease in Italy. The European Journal of Health Economics, 18(7), pp.847-858. UNEP (2013). Cost of inaction on the sound management of chemicals. Report Number: DTI/1551/GE. Vieira V M et al. (2013). Perfluorooctanoic acid exposure and cancer outcomes in a contaminated community: a geographic analysis. Environmental health perspectives, 121(3). Vierke L et al. (2012). Perfluorooctanic acid (PFOA) - main concerns and regulatory developments in Europe from an environmental point of view. Environmental Sciences Europe. v 24: (16). Viscusi W K and Masterman C J (2017). Income elasticities and global values of a statistical life. Journal of Benefit-Cost Analysis, 8(2), pp.226-250. VPR News, Bennington Homes Contaminated With PFOA Connect To Clean Municipal System. Wagner A (2018) Cape Fear River Watch, SELC sue Chemours for alleged GenX violations, StarNews Online, August 29 2018. Accessed 15.10.2018. Wagner A (2018). Chemours to pay $12 Million fine as part of GenX Agreement, StarNews Online, 21.11.2018. Accessed 01.12.2018. Wang Z et al. (2017). A never-ending story of per- and polyfluoroalkyl substances (PFASs)?. Environmental Science & Technology, Mar 7;51(5). The cost of inaction linked to PFAS exposure 141 Ward C, Hoosick Falls free bottled water program comes to an end., News10 (01 September 2017). Accessed 18.10.2018. Weber R, (2016). Presentation for Science and Policy of Organohalogens pre-Dioxin Symposium, accessed August 2018. WHO Europe (2016). Keeping our water clean: The case of water contamination in the Veneto Region, Italy. Winkens K et al. (2017). Early life exposure to per- and polyfluoroalkyl substances (PFASs): A critical review. Emerging Contaminants. 3: 55-68. doi:10.1016/j.emcon.2017.05.001. Website of G. Peters, Senator for Michigan (2018). Peters Provision Allowing Airports to Discontinue PFAS To Be Signed Into Law, Accessed 08.11.2018. Winquist A and Steenland K (2014). Modeled PFOA exposure and coronary artery disease, hypertension, and high cholesterol in community and worker cohorts. Environmental health perspectives, 122(12). Yamada T and PH Taylor (2003). Final Report: Laboratory scale thermal degradation of perfluorooctanyl sulfonate and related precursors. (Final Report for 3M Company). Yamada T et al. (2005). Thermal degradation of fluorotelomer treated articles and related materials. Chemosphere 61, 974-984. Zhou et al. (2018). Uncontrolled hypertension increases risk of all-cause and cardiovascular disease mortality in US adults: the NHANES III Linked Mortality Study. Scientific reports, 8(1), p.9418. 142 The cost of inaction linked to PFAS exposure Sammanfattning Hgfluorerade mnen (PFAS) kan skada mnniskors hlsa och miljn. Frutom att orsaka ett stort lidande fr drabbade individer medfr denna exponering stora kostnader fr samhllet. I den hr studien uppskattas kostnader fr samhllet som anvndningen av PFAS kan orsaka p grund av dess pverkan p miljn och mnniskors hlsa. En strre medvetenhet om kostnader och lngsiktiga problem som kan frknippas med anvndningen av hgfluorerade mnen kan hjlpa berrda myndigheter, beslutsfattare och allmnheten att ta beslut om effektiva och riskminskande tgrder. I studien granskas hur produktionen av PFAS, tillverkningen och anvndningen av produkter som innehller PFAS samt hur sluthanteringen av dessa har resulterat i omfattande miljfrorening och exponering av mnniskor. PFAS har hittats i miljn ver hela vrlden och nstan alla som lever i ett land med utvecklad ekonomi har ett eller flera PFAS-mnen i sin kropp. Eftersom de hgfluorerade mnena r s lnglivade i miljn kommer dessa mnen att finnas kvar p vr planet i hundratals, om inte tusentals r. Mnniskors och miljns exponering fr PFAS kommer att fortstta, och tgrder fr att minska denna exponering kommer medfra stora kostnader fr samhllet. Det r kostnader som till stor del ska betalas av offentliga myndigheter och skattebetalare. I rapporten studeras kostnader som samhllet kommer att behva betala i framtiden om tgrder inte genomfrs fr att begrnsa utslppen av PFAS. Studien fokuserar p de lnder som ingr i det Europeiska Ekonomiska Samarbetsomrdet (EES). "Cost of inaction" definieras som de kostnader som samhllet i framtiden kommer att behva betala om inga tgrder genomfrs idag fr att begrnsa utslppen av PFAS. De PFAS som omfattas i den hr studien r fluorerade ytaktiva mnen med en kolkedjelngd av C4-C14. Studiens mlsttning var att ta fram: 1. Ett ramverk fr att uppskatta samhllskostnader kopplade till negativa effekter p hlsa och milj i samband med exponering fr PFAS. 2. Monetra vrden fr dessa samhllskostnader, dokumenterade genom fallstudier. Slutsatser Fr att uppskatta de hlsorelaterade och miljmssiga kostnader fr samhllet utgick vi frn scenarier som bygger p antaganden, eftersom tillgngen p data r s begrnsad. Studien baserar sig p litteratur, information frn myndigheter och nyhetsartiklar. ven om studien och dess analys innebr vissa oskerhetsfaktorer finns det en relativt stor skerhet i en rad frgor: The cost of inaction linked to PFAS exposure 143 1. PFAS frekommer verallt i miljn och nstan alla mnniskor har PFAS i sina kroppar. Mtningar frn bde Sverige och USA visar att cirka tre procent av befolkningen exponeras fr halter som ligger ver de rekommenderade grnsvrdena, framfrallt via dricksvatten men andra kllor frekommer ocks. 2. PFAS kan hamna i miljn och i mnniskor p mnga olika stt. En klla till spridning av PFAS i miljn r via, till exempel anvndningen av brandslckningsskum fr brandbekmpning p flygplatser och p vissa industrianlggningar. PFAS sprids d till miljn och nr slutligen dricksvattnet. PFAS kan ocks spridas via anvndningen i konsumentvaror s som exempelvis pizzakartonger, klder och kosmetiska produkter. 3. Fr mnga av de omrden dr PFAS idag anvnds finns redan icke-fluorerade alternativ tillgngliga p marknaden. Drfr skulle vissa anvndningar redan idag kunna minskas. 4. Kostnaden fr att sanera mark och vatten uppskattas till mnga miljoner EUR. Den totala saneringskostnaden fr EUs medlemslnder uppskattas till minst hundratals miljoner EUR. 5. De negativa hlsoeffekter som kan bero p exponering fr PFAS r redan mnga och fortstter dessutom att ka i antal. Det finns ocks bevis fr att det kan uppst negativa effekter redan vid exponering fr relativt lga halter ven vid halter motsvarande lga (bakgrunds) niver. De rekommenderade grnsvrdena fr dricksvatten kan komma att snkas ytterligare nr mer information om risker fr hlsa och milj tas fram. Detta skulle ka kostnaderna fr den miljsanering som uppskattats i denna studie. Berkningarna i studien grundar sig p ett antal antaganden. Dessa antaganden har stmts av mot faktiska kostnadsuppgifter fr att skerstlla att de r kopplade till verkliga fall. Allteftersom mer information blir tillgnglig kommer berkningarna kunna bli mer exakta. I den hr studien har vi varit frsiktiga i vra antaganden, resultaten visar drfr p kostnader som sannolikt inte kan vara lgre, dremot skulle de kunna vara hgre. Det r troligt att kostnaderna kommer att ka i takt med att frhjda volymer och ett kat antal PFAS finns tillgngliga p marknaden. Att vnta nnu lngre med att begrnsa utslppen av PFAS kan f stora konsekvenser. Det kan innebra att antalet utslppskllor kar, att fler personer exponeras och att kostnaderna fr saneringen blir hgre. Froreningarna kommer ocks att sprida sig, vilket medfr att strre mngder mark och/eller grundvatten kommer att behva saneras. Metod Tv metoder har utvecklats under arbetet med denna rapport. Bda metoderna r baserade p fallstudier som berr exponeringen av PFAS. Den ena metoden anvnds fr att bedma hlsorelaterade kostnader. Den andra metoden anvnds fr att berkna kostnaderna fr miljsanering. Data specifik fr de nordiska lnderna har anvnts nr 144 The cost of inaction linked to PFAS exposure sdana har funnits, men uppskattningarna bygger ocks p information frn andra europeiska lnder, USA och Australien. Spridningsvgar (fallstudier) Fr att illustrera hur exponeringen av mnniskor och milj ser ut har fem fallstudier som fljer PFAS livscykeln (frn produktion och anvndning vid produkttillverkning till produkters anvndning och sluthantering) anvnts. Fr att f ytterligare uppgifter om direkta kostnader som uppsttt i samband med till exempel behovet att sanera frorenat dricksvatten har ven andra fall av PFAS-froreningar studerats. I fallstudie 1 studeras produktionen av PFAS i Europa och vilka froreningar denna har orsakat. Hr granskas froreningar som r kopplade till Chemours fabriker i Dordrecht (Nederlnderna); Mitenis anlggning i Veneto-regionen (Italien); och 3Ms anlggning nra Antwerpen (Belgien). Studien uppskattar att upp till 20 anlggningar aktivt producerar fluorokemikalier i Europa samt att dessa anlggningar r betydande kllor till utslpp av PFAS till miljn. Studien visar ocks att exponeringen fr arbetare vid dessa anlggningar r hg. Fallstudie 2 handlar om tillverkning samt anvndning av produkter som innehller PFAS. Industriella aktiviteter som kan slppa ut PFAS till miljn inkluderar textil- och ldertillverkning; metallpltering, inklusive krompltering; tillverkning av papper och pappersvaror; frger och lacker; rengringsprodukter; plast, hartser och gummi; samt biltvttar. I studien antas att ett intervall p 3-10 procent av dessa anlggningar anvnder PFAS. Ngra fluorokemiska produktionsanlggningar i de nordiska lnderna kunde inte identifieras. Dremot visar statistik frn Eurostat att annan industriell verksamhet som riskerar att slppa ut PFAS till miljn ger rum i regionen, ssom metallpltering och tillverkning av papper- och pappersvaror. I fallstudie 3 och 4 studeras anvndningen av kemiska produkter och varor som innehller PFAS. Vattenbaserade filmbildande skum (AFFF) som innehller PFAS har anvnts fr att slcka petroleumbaserade brnder samt i brandvningar, vilket har orsakat att grundvatten frorenats - srskilt kring flygplatser och militra baser. ven nrliggande samhllen har pverkats av frhjda niver av PFAS i dricksvattnet. I fallstudie 4 studeras anvndningen av varor och produkter som innehller PFAS s som mattor, livsmedelsfrpackningar och kosmetiska produkter. Dessa varor och produkter anvnds som exempel fr att visa hur anvndningen av en vara sannolikt kan leda till att mnniskor exponeras via livsmedel och absorption genom huden. Anvndningen leder ocks till utslpp av PFAS till miljn nr produkter skljs av eller rengrs och nr avloppsnt, reningsverk och s smningom recipient. I fallstudie 5 studeras effekterna frn sluthanteringen av produkter som innehller PFAS. Kommunal avfallsfrbrnning kan destruera PFAS i produkter vid driftstemperaturer p 1000C, men vid deponering kommer PFAS att finnas kvar ven efter att produktens krnmaterial brutits ner. Freningarna kommer s smningom frflytta sig till vtskor i deponin, och drefter till uppsamlingssystem fr lakvatten eller direkt till The cost of inaction linked to PFAS exposure 145 grundvatten och mark. De kan sedan n dricksvattentkter eller tas upp av vxter och bioackumuleras i livsmedelskedjan. Hlsorelaterade samhllskostnader Fr att berkna hlsokostnader fr samhllet har studien skt efter konsensus om hlsoeffekter kopplade till exponering fr PFAS. De vetenskapliga bevis som finns r motstridiga, men viss konsensus verkar rda vad gller leverskador, hga blodtrycksniver, nedsatt immunfrsvar (hgre infektionsrisk), kad risk fr skldkrtelsjukdom, nedsatt fertilitet, graviditetsinducerad hypertoni, preeklampsi, lgre fdelsevikt, samt testikel- och njurcancer. Metoden bygger p riskrelationer frn epidemiologiska studier fr populationer som i olika hg grad exponerats fr PFAS. Arbetare som exponerats fr PFAS p arbetsplatsen r exempel p en grupp med en hg exponeringsniv, medan samhllen som drabbats av PFAS p grund av nrheten till produktionsanlggningar, eller dr fluorerade AFFF anvnts, antas ha exponerats p medelniv. Denna exponeringsniv antas glla fr tre procent av den europeiska befolkningen, medan befolkningen i vrigt antas ha exponerats fr lga (bakgrunds-) niver. Tabell 42 ger en versikt ver de uppskattade rliga kostnaderna fr ett antal hlsoeffekter dr riskkvoter fanns tillgngliga fr berrda populationer. Exempelvis bedmdes den rliga hlsorelaterade kostnaden fr frhjd risk fr njurcancer p.g.a. yrkesmssig exponering fr PFASs uppg till mellan EUR 12,7 och EUR 41,4 miljoneri EES-lnderna.362 Den uppskattade hlsorelaterade kostnaden blev betydligt hgre fr svl frhjda som bakgrundsniver p.g.a. det stora antalet personer som d berrs. Den rliga hlsorelaterade kostnaden fr exponering vid tre olika niver av PFASs berknades till minst EUR 2,8-4,6 miljarder i de nordiska lnderna och EUR 52 till EUR 84 miljarder i EES-lnderna. Trots stora oskerheter och att berkningarna till stor del baseras p antaganden s tyder resultaten p att hlsokostnaderna fr PFASs-exponering r betydande. 362 P.g.a. ofullstndiga data gllande antal och lokalisering av produktionsanlggningar fr kemikalier och produkter innehllande kemikalier har de hlsorelaterade kostnaderna, orsakade av yrkesmssig exponering fr PFAS i de nordiska lnderna, inte kunnat berknas. 146 The cost of inaction linked to PFAS exposure Tabell 42: Uppskattade rliga hlsorelaterade kostnader vid exponering av PFAS Exonerings- Exponerad popula- Hlso endpoint niv tion & klla Yrkesmssig (hg) Frhjd (medium) Arbetare p kemiska produktionsanlggningar eller tillverkningsstllen Samhllen nra kemiska anlggningar etc. med PFAS i dricksvattnet Njurcancer Ddlighet (alla orsaker) Lg fdelsevikt Infektion Bakgrund (lg) Totalt Vuxna i den allmnna befolkningen (exponerade via konsumentprodukter, bakgrundsniv) Hypertoni Nordiska lnder Alla EEA lnder Exponerad population Ej tillgngligt rlig kostnad Ej tillgngligt Exponerad population 84000- 273000 rlig kostnad EUR 12,7- 41,4 miljoner 621000 8843 fdslar 45000 barn 10,3 miljoner EUR 2,1-2,4 miljarder 136 fdslar med lg fdelsevikt 84000 ytterligare dagar med feber EUR 0,7-2,2 miljarder 12,5 miljoner 156344 fdslar 785000 barn 207,8 miljoner EUR 41-9 miljarder 3354 fdslar med lg fdelsevikt 1,5 miljoner ytterligare dagar i feber EUR 10,7-35 miljarder Nordiska ln- EUR 2,8-4,6 der miljarder Alla EEA lnder EUR 52-84 miljarder Anm.: rligt antal berrda personer r det uppskattade antalet individer med en kad risk fr negativa hlsoeffekter p grund av olika exponeringsniver. En viss verlappning frekommer i uppgifterna ovan, eftersom arbetstagare och drabbade samhllen ocks utstts fr bakgrundsniver av PFAS. Samtidigt r dessa kostnader troligen underskattade p grund av bristen p epidemiologiska riskrelationer att anvnda fr berkning av andra hlsoeffekter och relaterade kostnader. Icke hlsorelaterade samhllskostnader (miljrelaterade) Den andra metoden innefattade att samla information om direkta kostnader som uppsttt i samhllen dr tgrder, ssom rening av dricksvatten, vidtagits fr att minska mnniskors exponering fr PFAS. Med dessa sammanlagda kostnader som grund berknades kostnadsintervall per person eller per fall. Dessa enhetskostnader anvndes sedan fr att rkna samman kostnaderna fr sanering nr miljexponeringen, till exempel koncentrationer i dricksvatten, verskrider vissa niver. Det br noteras att intervallen som anges i tabellen r stora, ven nr de normaliserats mot populationen. Avgrande fr om intervallen fr medelvrden kan berknas beror p vilken mngd data som finns tillgngliga. Nr det exempelvis gller kostnaderna fr att rena vatten fanns ett flertal estimat tillgngliga. I ett sdant fall r det osannolikt att det faktiska medelvrdet kommer att vara ett extremvrde i ngon av intervallets ndar som hmtats frn studierna. Det r drfr rimligt att minska det observerade intervallet till exempel genom att ta bort de uppskattningar som redan i tillrckligt stor omfattning exkluderats frn andra datakllor, det vill sga att de r att betrakta som avvikande vrden. Fr vissa kostnader finns dock mycket f uppskattningar tillgngliga. Var och en av de tillgngliga uppskattningarna kan vara lika gllande fr att ange medelvrdet. I ett sdant fall antas det observerade vrdeintervallet som ett intervall fr troliga medelvrden. The cost of inaction linked to PFAS exposure 147 I de fall d inget intervall finns att hmta frn kllorna i litteraturstudien har ett intervall uppskattats som till exempel intervallet +/- 90 % som anvnds fr att upprtta ett hlsovrdsprogram (definieras ej som en hlsokostnad i denna rapport d den avser hantering av problemet och inte effekter p mnniskors hlsa). I det hr exemplet r det angivna intervallet stort av tv anledningar; dels p grund av bristen p tillgngliga data samt p grund av den mjliga variationen i genomfrande av hlsovrdsprogram. Liksom fr de hlsorelaterade uppskattningarna har studien antagit att 3 % av den Europeiska befolkningen r exponerad fr dricksvatten med PFAS-halter ver de reglerade tgrdsniverna, vilket gr att de vattenreningsverk som frser denna del av befolkningen med dricksvatten kommer att behva frbttras och underhllas de nrmsta 20 ren. Antagandet om 20 r terspeglar mjligheten att saneringen ska kunna lsa problemen, kanske genom rening eller anvndning av alternativ eller mjligheten fr att saneringstgrder pgr under mnga r. P grund av de oskerheter som finns i analysen och tillgngliga data har kostnader fr sanering kvantifierats med hjlp av en scenariobaserad metod. Fr varje scenario har ett antal parametrar specificerats, exempelvis storleken av den drabbade befolkningen och tidstgngen fr underhllsarbete. Tabell 43 visar kostnadsintervallen fr olika kategorier av tgrder kopplade till arbetet med att terstlla och rena miljn. Tabell 43: Summerade kostnadsuppgifter fr icke-hlsorelaterade utgifter, ver 20 r, Vidtagen tgrd vid fynd av PFASs Enhet Bsta uppskattning Intervall frn studier Anvnt intervall vervakning av kontaminering frn industriell anvndning eller AFFF Kostnad per vattenprov Kostnad/fall av kontaminering EUR 340 EUR 50000 EUR 278-402 EUR 5200-5,8 miljoner EUR 278-402 EUR 25000- 500000 Hlsobedmning (inklusive Kostnad/person biovervakning) EUR 50 Inget intervall EUR 5-95 (+/-90 %) Total biovervakning & hlsobedmning per fall dr s ansetts lmpligt EUR 3,4 miljoner EUR 2,5 miljoner- 4,3 miljoner EUR 1 miljoner- 5 miljoner Temporrt tillhandahl- Kostnad/person lande av ofrorenad resurs Inga relevanta data Installering av ny pipeline Kostnad/person EUR 800 EUR 37-5000 EUR 100-1500 Uppgradering av vattenverk (kapital) Kostnad/person EUR 300 EUR 8-2200 EUR 18-600 Uppgradering av vattenverk (underhll) Kostnad/person EUR 19 EUR 8-30 EUR 8-30 Utgrvning och behandling av jord kontaminerad genom industriell eller AFFF anvndning Kostnad/kg PFASs Kostnad/fall EUR 280000 EUR 5 miljoner EUR 100000- 4,3 miljoner EUR 100000- 3 miljarder EUR 100000- 1 miljoner EUR 300000- 50 miljoner Tabell 44 visar kostnadsintervall fr olika tgrdskategorier relaterade till miljsanering fr de fem nordiska lnderna. Sammantaget rr det sig om ett kostnadsintervall p 46 miljoner-11 miljarder EUR. 148 The cost of inaction linked to PFAS exposure Tabell 44: Detaljerad versikt av intervall fr de Nordiska lnderna, med antagande att 5 % (bsta uppskattning 3 %) av populationen/befolkningen exponeras med halter ver gllande grnsvrden och att vattenrening behvs under en 20 rs period. Antal exponerade personer (3 %) Screening & vervakning Hlsobedmning Uppgradering av vattenverk & underhll Marksanering Totalt Danmark 170000 Finland 160000 Island 10000 Norge 160000 Sverige 290000 Norden totalt 790000 EUR 70000- 8,3 miljoner EUR 250000- 22 miljoner EUR 10000- 900000 EUR 170000- 20 miljoner EUR 480000- 47 miljoner EUR 280000- 27 miljoner EUR 270000- 26 miljoner EUR 20000-1,6 miljoner EUR 7,4 miljoner-274 miljoner EUR 7,2 miljoner-265 miljoner EUR 400000- 1,6 miljoner EUR 260000- 25 miljoner EUR 490000- 46 miljoner EUR 6,8 miljoner-250 miljoner EUR 13 miljoner-472 miljoner EUR 0-798 miljoner EUR 2,2 miljoner-2,1 miljarder EUR 100000- 86 miljoner EUR 1,6 miljoner-1,9 miljarder EUR 4,3 miljoner- 4,5 miljarder EUR 8 miljoner- 1,1 miljarder EUR 10 miljoner-2,4 miljarder EUR 1 miljoner-105 miljoner EUR 9 miljoner-2,2 miljarder EUR 18 miljoner- 5,1 miljarder EUR 46 miljoner- 11 miljarder Kostnadsberkningarna som anges i Tabell 44 r sannolikt mer robusta fr den sammanlagda europeiska nivn n fr den nationella nivn. Tabell 45 visar sammanlagda kostnader fr miljscreening, vervakning (dr froreningar har hittats), vattenrening, marksanering och hlsobedmning fr de fem nordiska lnderna samt fr vriga EES-lnder och Schweiz. Tabell 45: Sammanlagda kostnader fr miljscreening, vervakning nr froreningar upptckts, vattenrening, marksanering och hlsobedmningar Bsta uppskattning Lg Hg Danmark Finland Island Norge Sverige vriga EES+CH Totalt EUR 145 miljoner EUR 214 miljoner EUR 12 miljoner EUR 194 miljoner EUR 423 miljoner EUR 15,9 miljarder EUR 16,9 miljarder EUR 8 miljoner EUR 10 miljoner EUR 1 miljoner EUR 9 miljoner EUR 18 miljoner EUR 776 miljoner EUR 821 miljoner EUR 1,1 miljarder EUR 2,4 miljarder EUR 105 miljoner EUR 2,2 miljarder EUR 5,1 miljarder EUR 159,9 miljarder EUR 170,8 miljarder Motsvarande berkningar fr alla 31 EES-lnder och Schweiz ger ett kostnadsintervall fr miljsanering p EUR 821 miljoner till EUR 170 miljarder. De lgre och vre grnserna br betraktas som symboliska p grund av den begrnsade information som r tillgnglig. Baserat p informationen frn litteraturstudien finns det anledning att tro att de nedre grnserna kommer verskridas. En uppskattning i storleksordningen EUR 10-20 miljarder r skerligen rimlig. Troligtvis kan betydligt hgre utgifter n s komma ifrga om flera fall med miljardkostnader fr marksanering identifieras. En uppskattning av kostnaderna fr ett fall som identifierats i samband med en studie gllande staden Rastatt i The cost of inaction linked to PFAS exposure 149 Baden-Wurttemberg i Tyskland ligger inom intervallet EUR 1 till EUR 3 miljarder, ett intervall som antas komma att ka i omfattning med tiden. I detta fall anses kllan till utslppen ha varit frorenat material av tervunnet papper som spreds ut p jordbruksmark. Detta visar att allvarliga problem inte alltid r kopplade till flygflt och tillverkning av PFAS. Ett antal andra kostnader relaterade till frorening av PFAS ligger utanfr denna studies avgrnsning och kvantifiering. Dessa inkluderar minskat vrde p egendom, frsmrat rykte fr fretag som frorenar, ekologiska skador, samt kostnader fr att tgrda frorenade omrden som belastar berrda myndigheter - vilket inkluderar att informera allmnheten, att analysera frekomst av froreningar samt att vidta saneringstgrder. 150 The cost of inaction linked to PFAS exposure Annex 1: National / state screening for PFAS contamination This annex reviews the efforts of several EEA Member States and the USA to screen their territories to determine where PFAS contamination may pose risks to health and the environment. Two approaches for such screening can be distinguished. One approach is to identify the types of uses of PFAS that might lead to releases to the environment and then test water and soil in nearby areas for contamination. The other approach is to carry out comprehensive testing of drinking water supplies or wastewater treatment plant discharges and, if PFAS contamination is detected at levels of concern, to work upstream to identify the source of the contamination. Denmark Groundwater is the major source of drinking water in Denmark. Denmark therefore regularly monitors groundwater, under a program called GRUMO363, as well as the wells supplying waterworks. A 2014 screening investigation looked for the presence of 7 PFAS in groundwater where uses of PFAS may have led to contamination364 (Denmark has no direct production of PFAS). It drew on a 2005 survey of annual consumption of PFAS that estimated, of a total of 9-16 tonnes used that year, 50% was used in various textile, leather and paper products. PFOS and other PFAS were also used in paints and varnishes, in cleaning and polishing products, and in electroplating. Chromium plating was also targeted, since it was considered the largest consumer of PFOS in Europe at the time. Table A1.1 lists the industries and activities identified as potential sources of contamination, the numbers of screening investigations, and any PFAS contamination of groundwater identified in the 2014 screening report. 363 Geological Survey of Denmark and Greenland, Monitoring Programmes. 364 Danish Ministry of Environment (2014). Screeningsundersgelse af udvalgte PFAS-forbindelser som jor- og grundvandsforureng i forbindelse med punktkilder, Miljprojekt nr. 1600. The cost of inaction linked to PFAS exposure 151 Table A1.1: Number of sites identified as potential sources of contamination and overview Industry/activity No. sites investigated PFAS in groundwater? Concentration levels Fire training facilities 8 sites Yes Chromium plating 2 sites No Carpet industry 1 site Yes Paint industry 1 site No Landfills 4 sites No 2 sites ~ 100 ng/l 2 sites - >1000 ng/l 4 sites - none found na 1500 ng/l na na The investigation confirmed that fire drill sites were confirmed as potential sources of PFAS contamination, and recommended surveying the other large fire drill sites in Denmark that had used PFAS-containing aqueous film-forming foams (AFFFs). It reported on a comprehensive site investigation performed at another site (Copenhagen airport) that found concentrations of PFOS and PFOA at 500 times the Danish summed criteria of 100 ng/l (for 12 PFAS) in drinking water. The contamination on that site is contained by pumping and treatment by activated carbon filters to avoid contamination of drinking water. A 2016 study365 reviewed the Danish Product Register for information on professional uses of PFAS from 1983 to 2016. The reported data covered 152 compounds distributed over 27 industries and showed a decline of PFAS used from 98.5 tonnes in 2003 to 13.9 tonnes in 2016. It expanded the list of uses where contamination might have occurred to include the wood industry and furniture industry; the chemical industry, iron and metal industry and rubber and plastics industry; and other locations where direct emissions may have occurred, e.g., sites of chemical/oil fires. The study recommended that investigations for possible site contamination also should be carried out at sites where these types of uses occurred. Concerning drinking water, testing was carried out for a total of 446 samples from 318 waterworks wells from 2013 to 2018.366 The levels for all samples were below the summed criteria of 100 ng/l for the 12 PFAS monitored in Denmark. In 2017, the Danish Regions tested 1730 groundwater samples near potential sources of contamination for the content of the 12 PFAS in covered by the Danish drinking limit value.367 PFAS were found above the sum criteria in approximately 10% of the samples. The groundwater samples were taken in the upper part of the water table to identify soil contamination, rather than for testing of drinking water. 365 Danish EPA (2016). Kortlgning af brancher der anvender PFAS. 366 Personal communication; report not yet published (September 2018). Denmark has approximately 6,000 waterworks wells. 367 Danish Regions (2018). Forebyggelse & samarbejde. 152 The cost of inaction linked to PFAS exposure Sweden According to a 2016 Swedish EPA report368, 1,893 public water supplies were in use in 2015, 660 of which have been analysed for PFAS. About half of Sweden's public drinking water supply comes from surface water369; more than 80% of all surface water samples analysed contained detectable levels of PFAS. This includes sites with varying distances to identified PFAS sources. Available information concerning PFAS in groundwater is less reliable. Around 40% of the samples from areas with diffuse pollution, and 80% of samples from fire training sites contained detectable levels of PFAS. In 2014, a survey of local authorities controlling drinking water in Sweden carried out by Swedish National Food Administration showed that around 3.6 million Swedes drink PFAS-contaminated water370 - mainly the facilities using surface water. The levels were in most cases below 10 ng/l and were not considered to pose any threats to the public health. Also in 2014, five municipalities took actions related to their water supply in order to respond to PFAS contamination: Bstad, Ronneby, Halmstad, Uppsala and Botkyrka. Table A1.2 shows the number of sites sampled for different sources, and the analytical results. Table A1.2: PFAS detected in areas affected by different sources of emissions Source of PFAS No of sites sampled No of PFAS detected Average levels of PFAS detected (ng/l) Average levels of PFOS detected (ng/l) SW GW SW GW SW (10 PFAS1) GW (7 PFAS) Fire training site 214 271 18 - 3 193 43 927 Diffuse pollution 34 53 10 - 99,5 209,6 Waste facility 1 - - - Not analysed - for 10 PFAS Background area 60 - 14 - 13,5 - SW 1 207 42,3 21,6 1,5 GW 24 463 23,5 Note: SW = surface water. GW = groundwater. 1) PFOS, PFOA, PFHxS, PFHxA, PFHpA, PFNA, PFDA, PFBS, 6:2 FTSA and FOSA. The 2016 report confirmed that the presence of PFAS in water tends to increase with proximity to likely contaminated areas. Nevertheless, detectable levels are also found in background areas far from any identified sources. The same survey identified more than 2,000 potential sources from which PFAS might enter the environment; the most common types being WWTPs and major fire incidents.371 368 Swedish Environmental Protection Agency (2016). Highly fluorinated substances (PFAS) and pesticide, an overview of the presence in the environment (report in Swedish with English Summary). 369 Swedish Water & Wastewater Association (2016). Produktion av dricksvatten. 370 Swedish National Food Administration (2014). PFAA i rvatten och dricksvatten (report in Swedish) 371 IVL 2016 as stated in Swedish Environmental Protection Agency (2016) report. The cost of inaction linked to PFAS exposure 153 Table A1.3 lists the industries and activities identified as potential sources of contamination, the numbers of sites, and the estimated amount of PFAS released from each type of source into the environment. Table A1.3: Overview of various sources of PFAS and their estimated contributions to PFAS in the Swedish environment 20161 Industry/activity No. sites identified2 Estimated contribution of PFAS Production of PFAS Production of PFAS-containing fire foam Waste water treatment plants (WWTP) Landfills (emissions from other parts of the waste treatment process are unknown) Civilian airports (connected fire drilling sites) Military airports (connected fire drilling sites) Other fire drilling sites Fire incidents between 1998-2014 (registered by the Swedish Civil Contingencies Agency) Chromium plating industries Direct emissions from other industry Atmospherical deposition 0 2 336 365 10 18 295 9,000 Na 70 kg/year, of which PFOS represents 20 kg via water and 5 kg/year via sludge 70 kg/year (4 kg PFOS). 8 kg of total enters the environment; remainder enters WWTPs 1 600 kg total (380 kg PFOS) 1970-2000s 9 700 kg total (2 300 kg PFOS) 1970-2000s Unknown 660 kg total (150 kg PFOS) 11 (of which 3 use PFOS) Na Across all of Sweden 180 kg PFOS/year (2013, per exemption) via waste water and air Na 650-1 700 kg/year (25-30 kg PFOS) Note: Source: 2) A follow-up study in 2018 increased the number of fire incidents where AFFFs were used to 13,500 and doubled the number of other sites where AFFFs had been handled to approximately 8003 1) Hansson et al. (2016). IVL-report C 182: Sammanstllning av befintlig kunskap om froreningskllor till PFAS-mnen i svensk milj (report in Swedish). 3 Swedish EPA (2018). Frdjupad miljvervakning av hgfluorerade miljgifter (s.k. PFAS) och av vxtskyddsmedel i vatten NV-08978-16 (report in Swedish). Finland Finland has recognized the use of AFFFs as a major source of dispersal of PFAS compounds. The OECD chemical safety portal372 reports screenings of soil near airports and firefighting training centers along with screenings of drinking water.373 The first screening in 2015 focused on military sites and use of AFFFs. It collected samples from surface (6) and ground (18) water, wastewater (5), soil samples (4) and sediment samples (3). Half of the samples contained PFCs above detection levels, with most dominant compounds being PFOS and PFOA and to a lesser extent PFBA, PFHxA AND PFHpA. The samples taken from areas where firefighting foams were used showed the highest concentrations.374 372 OECD chemical safety portal (2018). Country information for Finland. 373 Ibid. 374 Ryynnen T (Construction Establishment of Finnish Defence Administration) (2017). Per- and polyfluorinated substances in the Finnish Defence Forces. 154 The cost of inaction linked to PFAS exposure Belgium (Flanders) Flanders has undertaken stand-alone screening in order to determine the scope of contamination from PFAS in groundwater and soil. Various risk locations were identified, based on types of activities carried out in their vicinity. The activities included firefighting sites, industrial uses of PFAS (chromium plating, textile and paper manufacturers and the paint industry). A total of 24 sites were selected and 40 soil and 1 groundwater sample were tested. In 66% of the risk locations, sample concentrations were 10 times higher than the reporting limit. In 24% of the locations, concentrations were 1000 times higher than the reporting limit.375 The USA (national monitoring) From 2013 to 2015, under the third Unregulated Contaminant Monitoring Rule (URCM 3), the US Environmental Protection Agency carried out testing for PFAS at a representative sample of public water systems serving less than or equal to 10,000 persons. Some 37,000 samples from 5000 public water systems were tested for PFOS, PFOA, PFNA, PFHxS, PFHpA and PFBS.376 A spatial analysis of the PFAS (PFOS and PFOA) found in drinking water due to the URCM 3 monitoring arrived at an estimate of six million U.S. residents served by public water supplies at levels above the US EPA's lifetime health advisory limit (70 ng/l).377 The number of industrial sites manufacturing or using PFAS, the number of military fire training areas, and the number of wastewater treatment plants were significant predictors of the concentrations of PFAS in the public water supplies tested. As of April 2018, at least 126 military bases in the USA reported potentially harmful levels of PFAS.378 A total of 401 active and base closure sites had at least one area where a known or suspected release of PFAS had occurred. Starting in 2017, the US Department of Defense (DoD) has undertaken a complete monitoring of its drinking systems throughout the country.379 As of 2018, 524 DoD water drinking systems were tested and 24 showed PFOS/PFOA levels above the US EPA lifetime health advisory levels. Monitoring of off-base private drinking systems resulted in the testing of 2,445 wells; 564 among those showed levels above the US EPA advisory levels.380 The DoD has already spent 200 million USD studying and testing water supplies and providing either filters, 375 OVAM (2018). Onderzoek naar Aanwezigheid van PFAS in Grondwater, Bodem en Waterbodem ter hoogte van Risicoactiviteiten in Vlaanderen. 376 EPA (2017). The Third Unregulated Contaminant Monitoring Rule (UCMR 3): Data Summary. 377 Hu, X et al. (2016). Detection of Poly- and Perfluoroalkyl Substances (PFASs) in U.S. Drinking Water Linked to Industrial Sites, Military Fire Training Areas, and Wastewater Treatment Plants, Environmental Science & Technology Letters (10)3, pp. 344-350. 378 Copp T, DoD: At least 126 bases report water contaminants linked to cancer, birth defects, Military Times, 26 April 2018. Accessed 01.11.2018. 379 US Department of Defense (2017). Aqueous Film Forming Foam Report to Congress. 380 US Department of Defense (2018) Addressing Perfluorooctane Sulfonate (PFOS) and Perfluorooctanoic Acid (PFOA). The cost of inaction linked to PFAS exposure 155 alternate wells or bottled water where supplies were contaminated. The cost of cleaning up PFAS-contaminated water at military sites was estimated at 2 billion USD. The US state of Michigan In 2017, the state of Michigan set up the Michigan PFAS Action Response Team (MPART) - a multi-agency group linking health, environment and other state government agencies in an effort to investigate possible hot spots of PFAS contamination and to act to protect drinking water sources.381 The program started after several incidents of contamination of water were discovered near military bases where AFFFs had been used as well as historic pollution at a former leather treatment factory. The state established a health standard for PFOS and PFOA (70 ppt, the same level as the US EPA lifetime advisory standard) and tested drinking water from public water supplies and in schools across the state. At the end of August 2018, the program had tested 892 of the state's 1,841 public water systems along with schools that operated their own wells). Four public water systems were found to have problems, including the city of Flint. The state's Department of Environmental Quality looked for the source of the contamination where concentrations exceeded background levels. Michigan is also testing effluent from 90 wastewater treatment plants, as part of a program monitoring industrial pre-treatment of wastewater. In addition, the Department of Environmental Quality is sampling for PFAS when investigating the state's Superfund sites. As of September 2018, the Michigan PFAS response website listed 34 sites where contamination had been identified.382 In April 2018, elevated levels of PFOS in the Flint River were linked to discharges from a plating and plastics company that had reportedly not used PFOS for years.383 A few months later a creek that fed into the Huron River was found to have 5,500 ppt of PFOS, more than 450 times the state limit for surface water. The contamination was traced to another company384 supplying chrome-plated plastic components to automobile manufacturers. At one time, total PFAS downstream on the Huron River tested at 56,868 ppt. Due to high levels of PFAS in fish caught in the Huron River, the state issued a "do not eat fish" advisory for 30 sq. miles of the river. 381 State of Michigan, Website dedicated to PFAS contamination. 382 State of Michigan, PFAS Sites Being Investigated. 383 No Author, Black Cloud of PFOS, The County Press, 15.04.2018.Accessed 08.11.2018. 384 Gardner P, Astronomical' PFAS level sets new Michigan contamination milestone, Mlive , 24.09.2018. Accessed 05.11.2018. 156 The cost of inaction linked to PFAS exposure Annex 2: Health impacts - additional information on evidence and calculations This annex presents additional information regarding the health assessments. Part 1 explains the calculations for each monetised health impact in more detail as well as the country-specific estimates. Part 2 presents an overview of the key epidemiological studies drawn upon for the assessment. It provides information about the region and population studied, the PFAS compounds in the contamination, the relevant health endpoint, the time period and the sample size. For example, the first study mentioned (Mastrantonio et al., 2017) was used for the calculation of health impacts among the population with elevated exposure through close residence to a chemical plant and AFFFs contamination. Upon request the excelspread sheets used for the monetarisation and valuation in this report can also be provided along with a guidance on how to use the estimation of costs for value transfer. Part 1: Estimating health impacts in exposed populations This section explains the calculations underpinning each of the quantitative estimates presented in Section 5.1. Occupational (high) exposure scenario The estimation for the occupational exposure scenario relied on a number of parameters and assumptions, which are listed below: assume that the number of PFAS-producing plants in the EEA ranges from 12 to 20 (Case Study 1); assume there are 352,764 small manufacturing plants and 780 large manufacturing plants across the EEA in the categories of manufacturing and commercial uses identified in Case Study 2 (textiles, leather, metal plating, paper and paper products, paints and varnishes, soaps and detergents, plastics and resins, car wash establishments); assume that the number of employees (exposed population) in small manufacturing plants is on average 30 while the number of employees in large plants is on average 300; The cost of inaction linked to PFAS exposure 157  assume that an estimated 3% to 10% of the types of manufacturing plants identified in Case Study 2 use PFAS-treated products; and the elevated risk of death due to kidney cancer from occupational exposure to PFOA (relative risk = 3.67). This estimate was obtained from an epidemiological study carried out in West Virginia.385 The elevated risk was assumed to apply to the exposed population in the Nordic and EEA countries. The calculation began with an estimation of the exposed population of workers in the Nordic countries and the EEA using information from the first three assumptions. Table A2.1 presents the results. Table A2.1: Estimating the size of the population with occupational exposure to PFAS Best estimate Assumption for calculation Ave. workers in each Nordic* EEA* Estimated exposed population PFAS plants 12-20 Manufacturing plants - small Manufacturing plants - large Total: 352,764 780 Assume upper es- 500 timate Assume that 3- 30 10% have PFAS treated products 300 - 10,000 31,749-105,829 702-2,340 334,508-1,091,692 Note: *The actual numbers are not known and therefore a best guess was made to support the calculation. We considered the death rate in this population which might be attributed to kidney cancer due to PFAS exposure. The number of deaths was estimated and compared with a scenario where occupational PFAS exposure was not present. This calculation relied on the third parameter from Steenland and Woskie, 2012. The elevated risk of kidney cancer mortality was evident for a quartile of the exposed population. Similarly, the calculation for this study assumed that the only a quarter of the population with occupational exposure faced a higher risk of kidney cancer due to PFAS exposure. Table A2.2 presents the calculation for the Nordic and EEA countries together. The lower bound is based on the assumption that 3% of selected manufacturing plants make PFAS-treated products while the upper bound uses the assumption of 10%. These assumptions were made due to the absence of actual figures. 385 Steenland K and Woskie S (2012). Cohort mortality study of workers exposed to perfluorooctanoic acid. American journal of epidemiology. 176(10) pp.909-917. 158 The cost of inaction linked to PFAS exposure Table A2.2: Calculation of monetised impact of elevated mortality from kidney cancer due to occupational PFAS exposure - EEA countries Exposed population Population that experiences elevated health risk Lower bound 334,508 83,627 Upper bound 1,091,692 272,923 General population death rate by kidney cancer Annual deaths in exposed population - baseline Not linked to PFAS exposure: Linked to PFAS exposure: Total value per year 0.01% 4.98 1.4 3.6 EUR 12.7 million 0.01% 16.27 4.4 11.8 EUR 41.4 million Explanation/source See Table A2.1 above Assumes only a quartile of the exposed population experiences an elevated risk of kidney cancer mortality Eurostat, 2015 standardised death rate Milieu calculations using the elevated risk information from (Steenland and Woskie, 2012). Milieu calculations using ECHA lower bound value of life (EUR 3.5 million) Elevated (medium) exposure scenario Three calculations were made for the scenario. The first was related to the costs of the elevated risk of all-cause mortality among adults. The second was an estimation of the number of births of low birth weight. The third was related to the higher risk of common childhood infections among children (immunotoxicity endpoint). Each is described below. The estimation for the elevated risk of all-cause mortality drew on two key sources of information. One was an epidemiological study from the Veneto region.386 Another was an estimate from Sweden on the human exposure to elevated levels of PFAS in groundwater. The key assumptions were: a relative risk of all-cause mortality due to elevated PFAS exposure. For this, we relied on an estimate from an epidemiological study carried out in the Veneto Region.387 The study found a relative risk of 1.11 and a 95% confidence interval of 1.10 to 1.12. The calculation assumes that the elevated risk found in the Veneto Region applies to the exposed population in the Nordic and EEA countries; up to 300,000 residents of Sweden (or about 3%) are exposed to source levels of PFAS above the limit value via municipal drinking water.388 The calculation assumes that the level of exposure in Sweden is equivalent across the Nordic countries and the EEA. 386 Mastrantonio M et al. (2017). Drinking water contamination from perfluoroalkyl substances (PFAS): an ecological mortality study in the Veneto Region, Italy. The European Journal of Public Health. Feb 1;28(1):180-185. 387 Ibid. 388 Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). The cost of inaction linked to PFAS exposure 159 Table A2.3 presents the calculation. The lower and upper bounds are based on the values of the 95% confidence interval from the estimate from Mastrantonio et al., 2017. Table A2.3: Calculation of annual monetised impact of elevated mortality due to elevated PFAS exposure - Nordic and EEA countries Nordic countries EEA countries Explanation/source Total population 20,698,030 Population with elevated exposure General mortality rate Annual deaths in exposed population - baseline Not linked to PFAS exposure: Linked to PFAS exposure: Total value per year 620,941 1.0% 6,458 Lower bound 5,871 587 EUR 2.1 billion 415,697,178 12,470,915 129,199 Upper bound 5,766 Lower bound 117,453 Upper bound 115,356 692 EUR 2.4 billion 11,745 EUR 41.1 billion 13,843 EUR 48.5 billion Population ages 19 years and up (Eurostat, 2017) 3% parameter estimate applied Eurostat, 2017 Milieu calculations using the elevated risk information from Mastrantonio et al., 2017. Milieu calculations using ECHA lower bound value of life (EUR 3.5 million) The estimation for the elevated risk of low birth weight drew on three sources of information: number of total live births and percentage of low birth weight by country (Eurostat); elevated risk of low birth weight due to PFAS exposure. The calculation assumed that an estimate from an epidemiological study could be extrapolated to the exposed population in the Nordic and EEA countries. The study found a relative risk of low birth weight of 1.50389; population exposure to elevated levels of PFAS, estimated to be 3%.390 This parameter was also used in the previous calculation. The calculation assumes that the level of exposure in Sweden is equivalent across the Nordic countries and the EEA. 389 Stein C R et al. (2009). Serum levels of perfluorooctanoic acid and perfluorooctane sulfonate and pregnancy outcome. American journal of epidemiology, 170(7), pp.837-846. 390 Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). 160 The cost of inaction linked to PFAS exposure Table A2.4 presents the calculation. Table A2.4: Calculation of births of low birth weight due to elevated PFAS exposure - Nordic and EEA countries Nordic countries EEA coun- Explanation/source tries Number of births per year Births with elevated exposure Low birth weight rate Annual low birth weight in exposed population - baseline Not linked to PFAS exposure: Linked to PFAS exposure: 294,777 8,843 4.6 % 407 271 136 5,211,464 156,344 6.8 % 10,631 7,088 3,544 Eurostat, 2016 3% parameter estimate applied Eurostat, 2017 Milieu calculations using the elevated risk information from Stein et al., 2009. The estimation for the elevated risk of fever is based on findings from a study from Denmark.391 The study found that children ages 1-4 years of age with the highest serum concentration of PFAS (10.19-25.10 ng/ml) had an increased risk of fever. This serum concentration level most closely corresponds with the elevated exposure scenario. The calculation assumed that all children aged 1-4 years with an elevated exposure would face an increased risk of fever. The key pieces of information and assumptions were as follows; number of children ages 1-4 years of age in the Nordic countries and the EEA (Eurostat); elevated risk of infection. The calculation relies on the aforementioned Danish study, which finds a relative risk of fever of 1.65.392 The calculation assumes that the elevated risk of additional days with fever found in the study from Denmark applies to the entire population at elevated exposure in the Nordic countries and the EEA; population exposure to elevated levels of PFAS, estimated to be 3%.393 This parameter was also used in the two previous calculations. The calculation assumes that the level of exposure in Sweden is equivalent across the Nordic countries and the EEA. 391 Dalsager L et al. (2016). Association between prenatal exposure to perfluorinated compounds and symptoms of infections at age 1-4years among 359 children in the Odense Child Cohort. Environment international, 96, pp.58-64. 392 Ibid. 393 Holmstrm et al. (2014). Nationell screening av perfluorerade froreningar (PFAA) i dricksvatten. Rapport no 2014/20 (In Swedish). The cost of inaction linked to PFAS exposure 161 Table A2.5 presents the calculation. Table A2.5: Calculation of additional days of fever among children 1-4 years old - Nordic and EEA countries Nordic countries EEA coun- Explanation/source tries Number of children ages 1-4 years Children with elevated exposure Fever days per year in baseline Fever days in exposed population Not linked to PFAS exposure: Linked to PFAS exposure: 1,507,631 45,229 4.7 days 212,576 128,834 83,742 26,159,812 Eurostat, 2016 784,794 3% parameter estimate applied 4.7 days Eurostat, 2017 3,688,533 2,235,475 Milieu calculations using the elevated risk information from Dalsager et al., 2016. 1,453,059 Background (low) exposure scenario The estimation of health impacts from background exposure to PFAS focused on the health endpoint of hypertension. It drew on epidemiological evidence gathered from the National Health and Nutrition Examination Survey in the United States, which focused on the adult population.394 The parameters and assumptions in the calculation include the following: the adult population in the Nordic countries and the EEA (Eurostat); the incidence rate of hypertension in the EU395; elevated risk of hypertension. The calculation relies on an estimate for the elevated risk of hypertension due to PFAS exposure.396 The study found an odds ratio of 1.63 and a 95% confidence interval of 1.2 to 2.2; elevated risk of death due to hypertension. The calculation relies on an estimate from Zhou et al., 2018. Min et al., 2012 find an increased risk of hypertension in half of the population. The calculation assumes that half of the exposed population in the Nordic countries and the EEA face an elevated risk of hypertension. Table A2.6 show figures from calculations of the monetised impact of elevated risk of hypertension due to background exposure to PFAS for the Nordic and EEA countries. 394 Min, J. Y. et al. (2012). Perfluorooctanoic acid exposure is associated with elevated homocysteine and hypertension in US adults. Occup Environ Med, 69(9):658-62. 395 European Heart Network (2017). European Cardiovascular Disease Statistics 2017. 396 Ibid. 162 The cost of inaction linked to PFAS exposure Table A2.6: Calculation of monetised impact of elevated risk of hypertension due to background exposure to PFAS - Nordic and EEA countries Nordic countries EEA countries Explanation/source Exposed population Population at higher risk Incidence rate of hypertension Number of new hypertension cases Elevated risk of hypertension due to PFAS exposure Not linked to PFAS exposure: Linked to PFAS exposure: Elevated risk of disease due to hypertension Number of deaths Not linked to PFAS exposure: Linked to PFAS exposure: 20,698,030 10,349,015 0.01 75,931 Lower bound 1.2 Upper bound 2.2 63,276 34,514 12,655 0.012 41,417 916 763 153 EUR 0.7 billion 416 500 EUR 2.2 billion 415,697,178 207,848,589 1,525,000 Lower bound 1.2 Upper bound 2.2 1,270,83 3 254,167 693,182 831,818 18,398 15,331 3,066 EUR 10.7 billion 8,363 10,035 EUR 35 billion Adult population (19 years and up), Eurostat, 2017 About half have an increased risk of hypertension (Min et al., 2012). European Cardiovascular Disease Statistics, 2017 Milieu calculations using the elevated risk information from Min et al., 2012. Zhou et al., 2018 Milieu calculations using the elevated risk information from Zhou et al., 2018. Milieu calculations using ECHA lower bound value of life (EUR 3.5 million) The cost of inaction linked to PFAS exposure 163 Part 2: Key epidemiological studies An overview of the epidemilogical studies reviewed in this report is presented in Table A2.7. Table A2.7: Overview of epidemiological studies reviewed Study Population PFAS Compound Health endpoint Time period Sample size Mastrantonio et al., 2017 Veneto Region PFOA, PFOS, PFHxS Mortality 1980-2013 Vieira et al., 2013 C8 Health Project PFOA Cancer 1996-2005 Steenland et al., 2010 Barry et al., 2013 C8 Health Project C8 Health Project PFOA, PFOS PFOA Steenland et al., 2009 Byrne et al., 2018 Min et al., 2012 Nelson et al., 2010 Bonefeld-Jorgensen et al., 2011 Simpson et al., 2013 Winquist and Steenland, 2014 Steenland et al., 2013 Bonefeld-Jorgensen et al., 2011 Steenland and Woskie, 2012 Brede et al., 2010 Jensen et al., 2015 C8 Health Project PFOA, PFOS St Lawrence Island, Alaska USA - background levels USA - background levels Greenland PFOA, PFOS, PFUnA, PFNA PFOA PFOA, PFOS, PFHxS, PFNA PFOA, PFOS C8 Health Project PFOA C8 Health Project PFOA C8 Health Project PFOA Greenland C8 Health Project workers Arnsberg, Germany Denmark PFOA, PFOS, PFHxS, PFNA, PFOSA PFOA PFOA PFOA, PFOS, PFNA, PFDA Uric acid Cancer Serum lipids Thyroid hormone Hypertension Cholesterol Breast cancer Stroke Hypertension and cholesterol Ulcerative colitis Breast cancer 2005-2006 2005-2006 2005-2006 2013-2014 2003-2006 2003-2004 2000-2003 2005-2006 2005-2006 2005-2011 1996-2002 Mortality None Miscarriage 1952-2008 2008 2010-2012 Shankar et al., 2012 Alexander et al., 2003 Bach et al., 2016 USA - background levels Decatur, Alabama (USA) Denmark Wang et al., 2016 Taiwan Yang et al., 2016 Beijing, China Ingelido et al., 2018 Veneto Region PFOA PFOS PFOA, PFOS, PFHxS, PFUnA, PFNA, PFHpS, PFDA PFOA, PFUnA, PFNA, PDFeA, PFUnDA, PFDoDA PFOA, PFOS, PFUnA, PFNA, PFDA PFOA, PFOS Cardiovascular disease Mortality; Bladder cancer Birth weight 1999-2003 1998 2008-2013 Fetal and post-natal growth 2000-2001 Thyroid hormones 2013 None 2015-2016 41,841 deaths; 143 605 residents in contaminated area; 588 012 residents in uncontaminated areas 29,118 cases of cancer; resident population over 500,000 54,951 adults 32,254 adults; 3,589 cancer cases 46,294 adults 85 adults 2,934 adults 860 individuals 31 breast cancer cases and 115 controls 32,254 individuals aged 12 years and above 32,254 individuals aged 12 years and above 32,254 individuals aged 12 years and above 250 breast control cases and 233 controls 1,308 workers; 243 deaths 138 individuals 2,874 in relatively exposed area and 336 in a relatively unexposed area 1, 216 adults 2,083 individuals; 145 deaths 1,507 mother-child dyads 223 mother-child dyads 157 mother-child dyads 507 subjects, 257 in high exposure areas 164 The cost of inaction linked to PFAS exposure Annex 3: Data used for environmentrelated cost calculations This annex presents the data used and the calculations carried out for the environmentrelated cost estimates in three parts: Part 1: The costs gathered via the case studies and additional research; Part 2: Data used in the aggregation of costs as presented in Section 5.2 of the study; Part 3: Full cost estimates by country. Upon request the excelspread sheets used for the monetarisation and valuation in this report can also be provided along with a guidance on how to use the estimation of costs for value transfer. Part 1: The costs gathered via the case studies and additional research This Annex reports the data collected for the following cost elements: 1. Monitoring to assess PFAS contamination where it is suspected (Table A3.2 and Table A3.3). 2. Provision of a temporary uncontaminated drinking water supply (Table A3.4). 3. Upgrading of water treatment works and ongoing costs for maintenance and re- placement and disposal of filters (Table A3.5). 4. Excavation and treatment of soils (Table A3.6). 5. Health assessments where contamination is found (Table A3.7). Exchange rates used All currency data have been converted to Euro in 2017 prices. The exchange rates used are based on Purchasing Power Parity (PPP), from OECD. Purchasing power parities are the rates of currency conversion that equalise the purchasing power of different currencies by eliminating the differences in price levels between countries. In their simplest form, PPPs show the ratio of prices in national currencies of the same good or service The cost of inaction linked to PFAS exposure 165 in different countries. PPPs are also calculated for groups of products and for each of the various levels of aggregation up to and including GDP. The basket of goods and services priced is a sample of all those that are a part of final expenditure: household consumption, government services, capital formation and net exports, covered by GDP. The original indicator is measured in terms of national currency per US dollar. It will be noted that there is variation between PPP rates and market exchange rates. Inflation has been accounted for using the HICP indicator from Eurostat. Key data for currency conversion to 2017 EUR are given in Table A3.1. All costs that follow in the tables, etc. presented in the report have been updated to 2017 EUR. Table A3.1: Conversion rates used EU28 HICP price index, EUR/Danish 2015=100 Krone 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 89.82 90.71 92.60 95.47 97.99 99.47 100.01 100.00 100.25 101.97 0.0995 0.0984 0.1008 0.1010 0.0999 0.0999 0.1006 0.1031 0.1000 0.1005 EUR/Norwegian Krone EUR/Swedish Krona 0.0892 0.0838 0.0836 0.0830 0.0836 0.0813 0.0794 0.0776 0.0727 0.0715 0.0900 0.0853 0.0848 0.0853 0.0873 0.0854 0.0845 0.0844 0.0811 0.0799 EUR/USD 0.7904 0.7609 0.7644 0.7543 0.7556 0.7345 0.7370 0.7551 0.7360 0.7274 EUR/GBP 1.1254 1.0718 1.0883 1.0680 1.0769 1.0561 1.0559 1.0832 1.0484 1.0196 Source: PPP conversion rates: OECD, https://data.oecd.org/conversion/purchasing-power-paritiesppp.htm Harmonised Index of Consumer Prices: Eurostat, https://ec.europa.eu/eurostat/web/hicp/data/database Monitoring to assess PFAS contamination where it is suspected Data identified so far for monitoring costs, in most cases relating to early site investigations, are shown in Table A3.2 and Table A3.3 for Europe and the USA. The annual sampling cost cited for Ronneby Airport is a figure for the period after site investigation and remediation, and is understood to possibly persist for many years. For most sites considered, the total costs of monitoring alone are in excess of EUR 100,000, and in several cases exceed EUR 1 million. Only partial estimates of cost have been identified for some sites. 166 The cost of inaction linked to PFAS exposure Table A3.2: Costs for PFAS monitoring, European data Location Sampling Cost Arlanda Airport Baden-Wurttemberg Bromma Airport Dusseldorf (fire at bottle depot) Kallinge Airport Kiruna Airport Norwegian coastline Oslo fjord Atmospheric contaminants (Norway) Screening of emerging contaminants (Norway) Riverine inputs and direct discharges to Norwegian coastal waters Contaminants in Norwegian lakes Pollutants in the Norwegian terrestrial and urban environment Ronneby Airport Stadtwerke Rastatt Stadtwerke Rastatt Ume Airport Uppsala Tyrifjorden Uppsala Sampling of groundwater, annual cost, to 2016 Annual cost for site investigations, 2017-2022 Investigation and research following identification of PFAS contamination Sampling and consultant costs for area where contamination was found Total cost of assessment EUR 5,200 EUR 8,000 EUR 6.1 million EUR 32,000 EUR 1.2 million Additional cost for water sampling within the monitoring programme Investigation of a confined potential soil contamination Estimated cost for implementing action plan, including: identifying risk objects, technical soil and water investigation, assessment of risks and need for measures, etc. Annual costs for analysis only, does not include sampling and reporting EUR 4,000 EUR 8,100 EUR 1.6 million EUR 36,000 EUR 21,000 EUR 32,000 EUR 36,000 EUR 4,000 EUR 14,000 EUR 28,000 Annual sampling cost after investigation and remediation Measuring groundwater wells in WSG Ottersdorf Measuring groundwater wells (WSG Rauental) Sampling and investigation Future annual additional costs to investigate potential PFAS-sources Field sampling Additional sampling of sediments, incl. reporting Admin, coordination fieldwork Analyses Analyses of additional sediment samples Reporting Budgeted future cost, fieldwork, analysis, reporting, admin Sampling associated with contaminated water supply Average cost for 2012-2014 Average cost for 2015-2017 EUR 3,400 EUR 440,000 EUR 110,000 EUR 110,000 EUR 16,000 to EUR 56,000 EUR 11,000 EUR 3,600 EUR 720 EUR 13,000 EUR 3,600 EUR 11,000 EUR 140,000 EUR 44,000 EUR 160,000 The cost of inaction linked to PFAS exposure 167 Table A3.3: Costs for PFAS monitoring, US data Sampling No. samples Washington State, USA US Military sites Monitoring Monitoring Monitoring Additional testing e.g. of specific incidents 500 8,000 24,000 Cost Cost/sample EUR 170,000 EUR 1,900,000 EUR 5,800,000 EUR 2,200,000 EUR 402 EUR 278 EUR 285 For the US data, the sampling regime under these different options includes costs of sample analysis and staff time to oversee testing and assist communities with drinking water contamination. Washington State foresees a need to monitor all public water supplies, down to Group B public water systems with fewer than 15 connections and fewer than 25 people per day. Provision of a temporary uncontaminated drinking water supply The only information obtained so far on the costs of supplying water temporarily is from the USA (Table A3.4). A widely reported figure of USD 10 million for provision of a temporary filtration system at Hoosick Falls, New York State and investigation into an alternative drinking water source is of a similar magnitude to the costs reported elsewhere for permanent systems. This suggests that the responsible authorities are seeking an alternative source with no contamination (as has been done at Bennington, just across the State border in Vermont), but that this will take several years to come online, essentially meaning that there is no difference between what is defined as a "temporary" solution for Hoosick Falls and permanent solutions elsewhere. Another figure is available for the rental of two large filtration tanks at Hoosick Falls at a cost of USD 300,000.397 It is unclear how this temporary installation differs to anything more permanent. A further temporary measure involved provision of free bottled water for collection from local supermarkets though there are no data available on the costs of this for Hoosick Falls. However, contamination at Peterson, Colorado, led to bottled water being provided. Costs for this are given as EUR 81,000, though it is unclear how many people were provided with the water which was intended specifically for residents using private wells or small drinking water systems.398 The total population in the area of concern is given as 60,000, of which it is estimated that 10-15,000 received water at levels above the health advisory threshold.399 Considering the likely costs of providing water per head, it seems likely that group served by the EUR 81,000 fund for bottled water was only a very small part of the 10-15,000 in the zone of highest concern. 397 DeMasi M, Hoosick Falls water filtration still weeks away, Albany Business Review, Jan 29, 2016. Accessed 05.08.2018. 398 KRRC (2016). Air Force Signs Contract for Bottled Water Distribution. 399 KRRC (2016). Health & Water Officials Try to Reassure Residents in Areas of PFC Contamination. 168 The cost of inaction linked to PFAS exposure Table A3.4: Costs normalised against population for temporary improvements to water systems Site Action Population Cost Cost/head Hoosick Falls Provision of bottled water Peterson, Colorado Provision of bottled water Temporary filtration system (1) Hoosick Falls Installation of temporary filters Temporary water filtration system (2) Hoosick Falls Temporary water filtration system and in- vestigation into an alternative drinking water source 3,400 3,400 3,400 No data EUR 81,000 EUR 270,000 EUR 79 EUR 8.6 million EUR 2,500 Press reports from Hoosick Falls indicate that when water supplies are declared clear of contamination there is still mistrust, with some people preferring to carry on drinking bottled water.400 Irrespective of the rationality behind the decisions made by individuals, this will also represent a social cost. Upgrading of water treatment works and ongoing costs for maintenance and replacement and disposal of filters Table A3.5 show costs normalised against population for improvements to water treatment in response to PFAS contamination, in cases where some reasonable estimate of the human population affected can be made. Cost data are also presented where no estimate of population is possible, for completeness. Within the table, categories of cost are grouped, starting with installation of advanced filters. Each group shows economies of scale, with cost/head reducing as the population served increases, as would be expected. A general problem with the cost data lies in understanding what elements of cost are covered. For example, the cost given for upgrading a WWTP may cover the capital costs of plant upgrade and subsequent maintenance costs, or the capital costs alone. In some cases costs are reported as being inclusive or exclusive of VAT, whilst in others no comment is provided. It is also often unclear whether maintenance costs, when provided, are represented as additional annual maintenance costs, or total maintenance costs. In some cases the cost data are separated out so that it is (reasonably) clear what is covered. In others, costs are reported as total but without further explanation. Where possible, original sources have been consulted in order to understand the data better. Estimates of the population served by each plant are approximate, based on the population of the nearest town or city. This introduces potentially significant error: for example, the low cost relative to size (EUR 8/person) for Dusseldorf may result from exaggeration of the population served. The issue is also highlighted in the case of Bennington (USA), for which two estimates of the cost per head for installation of new pipework to connect homes to an alternative water source are possible (not included in the table). The 400Ward C, Hoosick Falls free bottled water program comes to an end.News10 (01 September 2017). Accessed 18.10.2018. The cost of inaction linked to PFAS exposure 169 total cost of this action is given as USD 20 million401 (EUR 17 million). To derive a low estimate of damage per head, we divide by the total population of Bennington, 15,764, and reach a figure of EUR 1,085 per person. However, Saint-Gobain, the company funding the work as it is linked to a factory that they were responsible for, state that they have provided connection to 200 homes. Assuming an average occupancy rate of 2.5 people per home gives a total population 500. Dividing the USD 20 million cost of the work by this much lower population drives cost per head up to EUR 34,000 (and per property to EUR 85,000). With respect to variation in estimates, a Swedish expert who has worked on PFAS issues at a number of airports402 risk assessment and investigating measures might vary between SEK 5-15 million, remediating individual water bodies for drinking water abstraction could cost between SEK 0-15 million, and other measures to remediate the site could be estimated to SEK 20-50 million. Likewise, Avinor in Norway estimated that remediation might cost between NOK 3-30 million per airport.403 Accepting these uncertainties, the dataset provides reasonable ranges for feeding into the analysis. As before, the dataset may be improved by researching the costs of improved water treatment without specific reference to PFAS. An issue for aggregation of results concerns the size of communities affected. Clearly, not all will be very small or very large or average. This may best be addressed through sensitivity analysis. 401 VPR News, Bennington Homes Contaminated With PFOA Connect To Clean Municipal System. 402 Personal communication from Niklas Lwegren, Project leader contaminated sites, Swedish Transport Agency. 403 Heggelund, A.(2017) Norwegian EPA, personal communication September 2018. 170 The cost of inaction linked to PFAS exposure Table A3.5: Costs for upgrading water treatment works, by activity Incident Description Population 1 Cost Cost/person Installation of filters Europe Arlanda Dusseldorf Estimated cost of new treatment plant for surface water leaving site preventing contamination of Lake Malaren, used for drinking water Total cost of managing PFC in the contaminated drinking water Costs of remediation after AFFF use 1,200,000 Total costs of remediation Kallinge(SE) Landvetter Malmo Stadtwerke Rastatt Veneto USA Brunswick County Cape Fear Cape Fear River Cape Fear River Hoosick Falls (NY) Issaquah Moose Creek Peterson Air Force base Tennessee River (Decatur) Warrington Installation of new pipelines, filters Operation (2016) and installation of new (2017) treatment plant for surface water Establishing water treatment facility and analysing water Cost of treatment facility, sampling and dike cleansing Estimated cost for future remediation and monitoring in surrounding recipients Rauental water works 4,561 Installation of filters 120,000 Upgrade treatment plant with activated carbon 107,000 Upgrade treatment plant with granular activated carbon (Cape Fear) Upgrading of treatment plant installation of reverse-osmosis system (Brunswick County) Expansion of additional treatment works 360,000 107,000 "Temporary" water filtration system 3,400 Installation of water treatment systems Installation of granular activated carbon system Contamination from AFFFs, installation of water treatment systems for known contaminated wells Installation of carbon filtration systems 30,234 9,000 60,000 55,000 Installation of carbon filtration systems 23,000 Temporary water filtration system EUR 180,000 >EUR 13,000,000 >EUR 12,000,000 >EUR 100,000,000 EUR 4,300,000 EUR 250,000 >EUR 10 > EUR 8 > EUR 80 EUR 943 EUR 140,000 EUR 110,000 EUR 100,000 EUR4,000,000 EUR 2,100,000 EUR 18 EUR 72,000,000 EUR 33,000,000 EUR 72,000,000 EUR 28,000,000 EUR 7,400,000 EUR 673 EUR 92 EUR 672 EUR 2,200 EUR 880,000 EUR 2,700,000 EUR 3,300,000 EUR 29 EUR 300 EUR 55 EUR 3,700,000 EUR 67 EUR 12,000,000 EUR 522 USA Hoosick Falls Temporary water filtration system Annual maintenance cost of water treatment works 3,400 EUR 270,000 EUR 79 Europe Landvetter Stadtwerke Rastatt Annual operating cost Rauental water works 25,000 (50% of Rastatt) EUR 40,000 EUR 750,000 EUR 30 The cost of inaction linked to PFAS exposure 171 Incident Uppsaala Description Annual cost for activated charcoal filtration, 2012 2014-2017 average Veneto Annual operating cost USA Moose Creek Tennessee River (Decatur) Annual maintenance cost Annual maintenance cost Rebuilding of water treatment works Population 1 120,000 Cost EUR 30,000 EUR 110,000 EUR 920,000 Cost/person EUR 8 9,000 55,000 EUR 230,000 EUR 550,000 EUR 26 EUR 10 Europe Stadtwerke Rastatt Uppsaala Ottersdorf water works Reconstruction of treatment plant to deal with PFAS Install new pipelines / connections 25,000 (50% of Rastatt) EUR 3,900,000 EUR 260,000 Europe Jersey 2 Jersey 2 Kallinge Kallinge Stadtwerke Rastatt Veneto Connection to water supplies Jersey Water Mains connection costs Installation of new pipelines, filters Costs of providing alternative water supply via new pipe connections between 2013 and 2015 and use of a new set of carbon filters Water pipe between water works at Muggensturm and Rauental Water pipe Rauental-Lochfeldstr Installation of new pipelines (not carried out) 161 161 4,561 4,561 49,100 120,000 Other costs EUR 210,000 EUR 810,000 EUR 6,650,000 EUR 4,300,000 EUR 1,306 EUR 5,037 EUR 1,458 EUR 943 EUR 900,000 EUR 910,000 EUR 61,700,000 EUR37 accounting for both pipelines EUR 514 Europe Buncefield UK Lost opportunity cost from closure of a borehole Jersey 2 Site investigation 161 Jersey 2 Remedial works to old fire ground 161 Jersey 2 Fees, etc. 161 Jersey 2 Capital works on Fire Training Ground 161 Uppsaala Risk analyses and planning of measures to safeguard Uppsala's drinking water from pollution (not only PFAS) Time spent within the public water authority (2012- 2017, average annual cost) EUR 2,600,000 EUR 1,430,000 EUR 450,000 EUR 910,000 EUR 6,300,000 EUR 320,000 EUR 8,893 EUR 2,799 EUR 5,659 EUR 39,179 EUR 55,000 Note: 1)Population estimates provided here are for the municipalities identified in the case studies and hence are not necessarily specific to the number of people served by water treatment works. There may be significant uncertainty in these figures, especially for the largest towns and cities included. 2)The Jersey Airport case provides a breakdown of cost elements, as shown. It is possible that these costs are included in other estimates, but without disaggregation. 172 The cost of inaction linked to PFAS exposure Excavation and treatment of soils Costs relating to environmental remediation, mostly in relation to soil cleaning but including some groundwater cleaning also, are presented in Table A3.6. Table A3.6: Costs related to remediation of environment, after contamination from PFAS due the different sources Incident Year1 Description Cost (EUR)2 European cases Arlanda Airport (SE) 2016 Future Clean up of agricultural fields in Baden-Wurttemberg (DE) Bromma Airport (SE) Copenhagen Airport Future Future Future 2016 Contamination due to Dusseldorf Airport (DE) Population = 1.2 million urban area Contamination around Jersey Airport (UK) Population affected = 67 properties Contamination around Nurnberg Airport (DE)3 Contamination around Oslo Airport (NO)4 2014 1993 NA NA Contaminated soils, Schiphol (NL) Visby Airport (SE) 2008 Future US cases Minnesota contamination due to 3M factory disposal sites (US) 5 2002 Contamination in Warrington due to use of AFFFs (US) 2016 Cost of testing soil clean-up Est. cost of pilot study on remediation costs Estimated costs of remediation Estimate of changing contaminated soil EUR 7,600 EUR 80,000 to EUR 400,000 EUR 800,000 Up to EUR 3 billion Estimated future remediation costs Clean-up of site and reconstruction of fire training area Cost of 3 wells controlling the point sources: Estimated total remediation cost, up to: EUR 340,000 EUR 15 million EUR 2 million EUR 100 million Estimated total remediation cost 7.08 million Initial budget set by Nurnberg Airport for PFAS remediation Removal of 0.6 kg PFAS from stony area Removal of 0.5 kg PFOS/year by treatment facility at fire drill sites Removal of 50,000 m3 of soil, 143 kg PFOS EUR 10,000,000 EUR 1.9 million EUR 2.2 million EUR 30-40 million Estimated future remediation costs EUR 800,000 to EUR 1.4 million Total cost for 10 years for treatment of surface/ground water, sediment and soil at 3 sites Estimation of total costs for environmental restoration EUR 36 million EUR 77.7 million Note: 1) Year might refer to year of detection, or the year costs were incurred. 2) Costs in other currencies are converted to Euro, using average annual rates for the year they incurred. 3) Weber R, (2016). Presentation for Science and Policy of Organohalogens pre-Dioxin Symposium, accessed August 2018. 4) Norwegian Environment Agency (2016). PFAS-forurensning i grunnen Oppsummering fra workshop 26. Rapport M-622. 5) Legal settlement presented in table 19. The cost of inaction linked to PFAS exposure 173 The costs of dealing specifically with contaminated soils reported in Table A3.6 equate to the following estimates per unit of material (for those sites where data on both costs and quantity of PFAS were available): Nurnberg Airport: Schiphol Airport: Oslo Airport: EUR 100,000/kg mixed PFOS, PFHxS, PFBS EUR 200,000-280,000/kg PFOS; EUR 600-800/m3 soil EUR 2.1 to EUR4.3 million/kg PFOS Variability in the range is not surprising given that the costs for soil remediation are a function of several factors, including404: the quantity of PFAS that was spilled or emitted; the presence of other contaminants that need to be eliminated; the quantity of soil that has been contaminated; the type of soil and its qualities for retaining PFAS; variability in the use of sites and surrounding lands and waterbodies that will influence the desired level of remediation. Three Norwegian airports have modelled the costs of PFAS removal (including both water and soil), using different combinations of methods and different levels of allowed remaining concentrations. For Kristiansand airport, the figures range from around NOK 29.5 to NOK 332.5 million (EUR 2,1-24 million); from NOK 6.3 to NOK 91.3 million (EUR 0.5-7.1 million) for Harstad/Narvik; and from NOK 5.7 to NOK 113.4 million (EUR 0.4-8.1 million) for Svalbard Longyearbyen. Due to the highly hypothetical nature of these cost figures, as well as the vast number of cost estimates generated by the various choices of method and target concentration, they have not been included into the cost tables presented above. For the full details of the cost of the various remediation scenarios, the reader should consult the original reports.405 Health assessments where contamination is found A final category of cost concerns health assessment of the population in cases where contamination above permitted levels is identified. This category of cost is accounted for here, rather than in the section on "health costs", as associated costs relate to management of the problem rather than the health or environmental damage caused. 404 National Research Council (1997). Innovations in Ground Water and Soil Cleanup: From Concept to Commercialization. Chapter 6. National Academy Press. Chapter 6. 405 Avinor (2018). PFOS I Focus, (In Norwegian) Accessed 05.09.2018. 174 The cost of inaction linked to PFAS exposure For the Veneto Region, WHO provides the following information on a human biomonitoring study of the affected population: Dates: From July 2015 to April 2016 Population: Exposed and unexposed population Lead organisations: National Institute of Health, the Regional Environmental Protection Agency and health care trusts in the areas identified as being most affected Also involved: six local health and social-care units and 14 municipalities (seven exposed and seven unexposed). Sampling: Serum samples were taken from 507 people aged 20-49 years. The participants were also requested to complete a questionnaire on their dietary habits, water-supply sources and consumption of local food. Because of the contribution of other factors to PFAS body burden, the biomonitoring study also included a subgroup of 120 people living and working in agricultural areas, or working with livestock. Separately, WHO describes a Health Surveillance Plan: Dates: Started in December 2016. Population: The Plan covers five local health units and involves almost 85 000 people between the ages of 14 and 65 years Objectives: To identify areas of expected/possible health impact, using data on PFAS contamination of the water supply before the installation of filters. Screening activities: Biennial screening of the exposed population for cancer was introduced, starting with 14 year-olds in December 2016. The reason for choosing youth to begin with was that unhealthy lifestyles are not associated with this age group; thus, if high PFAS concentrations and/or significant metabolic changes were found, they could provide an insight into the correlation between exposure to PFAS and health outcomes. People with unhealthy lifestyles are informed of the risks to their health and provided with support in modifying their behaviour. Those with PFAS serum concentrations higher than the median for the Italian population, and/or showing biochemical or blood-pressure changes, are taken over by their family doctors and placed on a second-level care path for the timely diagnosis of diseases related to PFAS exposure. The Veneto Region has a regional PFAS screening information system, which manages the entire survey process, from the mailing of invitation letters to the delivery of the results and the development of the most representative health indicators. The programme is completely free of charge for the target population. An ad hoc surveillance plan is scheduled for pregnant women and those working in the manufacture of these substances. The cost of inaction linked to PFAS exposure 175 A series of further studies were initiated: ecological study of pregnancy and birth outcomes; occupational retrospective cohort study of employees at the chemical plant; retrospective ecological study of the exposed population considering mortality and morbidity data in the region over the period 2007-2014; retrospective ecological studies on cancer incidence over the period 1997-2013. Costs were reported above as EUR 4.3 million, equivalent to EUR 50 per person covered by the plan. WHO reports that this covers only the first two years of the Health Surveillance Plan. Biomonitoring has been carried out around Ronneby Airport in Sweden since 2014, following identification of PFAS contamination. The following costs have been identified406 (see Table A3.7): Table A3.7: Biomonitoring costs related to PFAS contamination around Ronneby Airport, 2014-2018. Year Activity Cost 2014-2015 2014-2015 2014 2015 2016 2017 2018 2014-2018 Risk assessment, risk communication, advice related to bio-monitoring study PFAS analyses related to bio-monitoring study Bio-monitoring study (not more specified) Bio-monitoring study (not more specified) Bio-monitoring study (not more specified) Bio-monitoring study (not more specified) Bio-monitoring study (not more specified) Average annual cost of biomonitoring study (including analyses) EUR 52,000 EUR 170,000 EUR 420,000 EUR 460,000 EUR 700,000 EUR 450,000 EUR 340,000 EUR 510,000 Part 2: Data used in the aggregation of costs Additional data have been used in the aggregation of costs presented in Section 4.2 of the study, as follows: Population (Table A3.8) Water consumption (Table A3.9 ) Number of wastewater treatment plant (WWTP) (Table A3.10) Number of plant or sources providing drinking water (Table A3.11) Number of airports (Table A3.12) Number of landfill and incineration sites (Table A3.13) Information covers the EU28, Norway and Iceland to the extent that data are available. Data for the USA are also included for reference, given that much of the information used in this report is of US origin. 406 Personal communication, K. Jakobsson, Oct 2018. 176 The cost of inaction linked to PFAS exposure Much of the analysis is based on extrapolation against estimates of the population affected. The population data provided in Table A3.8 are taken from Eurostat and the median projections under the UN's World Population Prospects. Some countries show a significant increase in population over the coming years (e.g. Denmark, France, Iceland, Luxembourg, Norway and Sweden), others show little change (e.g. Czechia, Finland, Germany, Italy, Slovakia) and some others a fall in population (e.g. Bulgaria, Greece, Lithuania). These changes will affect any quantification based on extrapolation of existing data on population. Table A3.8: Population, 2015-2050. Source: Eurostat for EU28, UN median projections for others 2015 2020 2030 2040 2050 Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Switzerland Sweden UK EU28 8,576,261 11,208,986 7,202,198 4,225,316 847,008 10,538,275 5,659,715 1,313,271 5,471,753 66,415,161 81,197,537 10,858,018 9,855,571 336,728 4,628,949 60,795,612 1,986,096 2,921,262 562,958 429,344 16,900,726 5,166,493 38,005,614 10,374,822 19,870,647 5,421,349 2,062,874 46,449,565 8,238,610 9,747,355 64,875,165 508,401,408 9,005,487 11,580,268 6,954,254 4,091,559 869,041 10,652,407 5,887,449 1,317,940 5,561,792 67,818,978 83,751,689 10,560,467 9,789,630 354,222 4,852,123 60,718,572 1,911,668 2,749,762 628,950 452,542 17,410,756 5,403,704 37,930,818 10,209,628 19,259,049 5,458,718 2,075,778 46,562,044 8,647,547 10,293,412 67,236,507 515,591,288 9,675,572 12,264,124 6,408,361 3,954,893 919,997 10,691,890 6,298,421 1,306,181 5,697,608 70,525,154 84,613,298 9,944,658 9,665,170 383,538 5,146,475 60,350,475 1,743,960 2,410,874 754,522 488,632 18,393,443 5,878,930 37,213,790 9,880,173 18,023,954 5,464,199 2,080,145 47,110,106 9,477,452 11,237,236 71,563,991 523,827,302 10,087,623 12,844,259 5,933,535 3,819,863 954,320 10,552,301 6,564,333 1,283,732 5,722,378 72,915,525 84,133,642 9,419,973 9,471,313 403,548 5,396,380 59,982,002 1,598,786 2,128,883 860,808 505,921 19,035,643 6,268,216 35,840,028 9,553,608 17,069,777 5,373,043 2,066,086 48,244,792 10,234,794 11,994,364 75,004,352 528,357,270 10,247,691 13,273,155 5,564,146 3,674,791 984,402 10,478,190 6,685,016 1,256,975 5,687,527 74,376,832 82,686,973 8,918,545 9,287,196 415,151 5,693,430 58,968,137 1,506,055 1,957,377 938,416 513,081 19,235,467 6,568,489 34,372,849 9,116,350 16,331,359 5,261,609 2,045,090 49,257,477 10,977,129 12,681,084 77,568,588 528,567,808 USA 325,127,634 337,983,029 362,628,830 383,165,322 400,853,042 Data on total water consumption are shown in Table A3.9 covering not only drinking water but also industrial, agricultural and commercial uses. Overall, most supplies (80%) are taken from surface water. However, data demonstrate significant variation between countries with respect to the reliance on ground water, with Denmark and Malta obtaining more than 90% of their water from groundwater, whilst Bulgaria, Romania and Finland take more than 90% of their water from surface sources. The cost of inaction linked to PFAS exposure 177 Table A3.9: Water consumption by country and source, million m3 Million m Surface Ground Surface + ground Total gross abstraction Austria Belgium Bulgaria Croatia Cyprus Czech Republic Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland United Kingdom 4,480 5,071 225 82 1,237 9 1,525 6,298 24,400 27,195 4,297 4,516 561 92 254 20 3 8,465 8,486 5,868 248 714 26,613 2,342 1,000 5,232 632 558 428 150 366 737 199 264 5,608 5,841 5,611 492 196 155 157 26 43 1,016 2,608 590 326 182 6,304 348 1,005 2,053 5,082 5,629 653 232 1,603 746 1,724 6,562 30,008 33,036 9,908 5,051 3,011 757 248 411 45 45 9,482 11,094 6,458 574 895 32,916 2,690 2,005 7,285 4,480 5,071 225 82 1,237 9 1,525 6,298 24,400 27,195 4,297 4,516 561 92 254 20 3 8,465 8,486 5,868 248 714 26,613 2,342 1,000 5,232 % surface 88% 90% 34% 35% 77% 1% 88% 96% 81% 82% 43% 89% 74% 37% 62% 43% 6% 89% 76% 91% 43% 80% 81% 87% 50% 72% % ground 12% 10% 66% 65% 23% 99% 12% 4% 19% 18% 57% 10% 26% 63% 38% 57% 94% 11% 24% 9% 57% 20% 19% 13% 50% 28% Table A3.10 shows the number of waste water treatment plants in the EU28, Iceland and Norway. Data are taken from Eurostat, and show information for the latest year for which data are available for each country. In most cases data are taken from the period 2010-2014. Older data are highlighted in red. Data for "urban" and "other" sites have been combined. The columns indicate different levels of treatment, as follows: T1: Primary treatment only, removing solid material T2: Secondary treatment, as T1 but also digesting dissolved and suspended organic materials, sometimes with disinfection to kill pathogenic bacteria T3: Tertiary treatment, as T2, but with a "polishing treatment" such as the use of microfiltration or synthetic membranes to further purify the water T3N: T3 plant with additional nitrogen removal (included in T3 total, and will include some plant also with phosphorus removal) T3P: T3 plant with additional phosphorus removal (included in T3 total, and will include some plant also with nitrogen removal). 178 The cost of inaction linked to PFAS exposure Table A3.10: Number of waste water treatment plant, category other + urban, in the EU28, Iceland and Norway Total T1 T2 T3 (total) T3N T3P Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden UK Total 1,842 1,222 90 316 191 2,636 906 1,044 202 3,275 12,590 235 739 18 1,063 2,717 1,165 717 251 4 555 2,240 4,296 4,287 826 568 352 2,041 1,243 8,047 55,678 412 10 195 50 177 216 23 2,307 10 13 217 178 207 78 130 1 4 1,565 363 219 125 4 35 740 7,279 791 485 54 109 133 1,314 273 539 647 4,824 35 278 13 536 510 640 578 98 3 158 83 2,970 1,617 518 346 314 974 325 5,151 24,316 1,051 325 26 12 88 1,272 456 289 202 2,605 4,028 200 451 310 1,876 48 61 24 393 592 963 116 89 97 36 1,032 918 2,156 19,716 806 255 26 15 7 546 311 98 50 2,516 3,540 198 338 30 1,345 11 56 2 3 332 47 84 32 722 132 11,502 889 251 22 1 6 83 446 231 202 1,719 3,112 130 369 215 838 56 1 311 27 34 24 509 960 10,436 Note: Italics used where latest data are from before 2010. Blank cells: no data or zero Source: Eurostat Database, Wastewater treatment plants by treatment level [env_ww_plt]. Numbers of waters supply zones are shown in Table A3.11. It seems likely that the use of filters of some kind would be applied to larger sites, whilst for smaller sites it may be more economical to provide water from alternative sources. Significant variation is seen between countries. For example, in the Netherlands very few people are served by small supplies or very small supplies, whilst these serve around 50% of the population in Lithuania. The cost of inaction linked to PFAS exposure 179 Table A3.11: Number of water supply zones in the EU28, Iceland and Norway* Total number of supply zones Large zones (>1,000 m3/d), >5,000 persons Small zones (10-1,000 m3/d), 505,000 persons Small zones: number of consumers Very small zones (<10m3/d), <50 persons Very small zones: number of consumers Austria Belgium Bulgaria Croatia Cyprus Czech Republic Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK 8,708 1,113 42 10 3,001 1,220 1,359 29,101 6,959 830 300 6,900 2,000 2,044 118 22 185 26,710 3,356 2,910 2,872 1,080 5,552 4,300 2,914 208 211 159 30 208 20 2 10 284 1,433 2 8,500 1,750,000 31,000 + 170,000 750,000 2,790 1,300 1,630,000 >500 public + 500,000 private wells 70,000 1.1 million to 3 million at weekends 71,300 1,200 900,000 thousands5 20,500 16,500,000 7,100 120,000 ~500,000 people (0.7%) are not served by centralised supply3 800 1,600,000 1,750 3,300,000 50 1,500 2,275 560,000 200,000 700,000 497 1,836 98 20 175 142,095 500,000 180,000 430,000 80,000 19 300,000 25 0 0 2,373 900,000 400 0 0 10,815 2,068 4,000 11,074 4,500,000 8,956 1,000 11,000 796 364,471 800k private wells 4 1.2 million permanent 266 1 222,488 1 24,000 1 95,000 1 Note: Cells in Italics calculated by subtraction of small from large zones where appropriate. 1) UK data for small and very small zones only available for Northern Ireland and Scotland. 2) Source: https://www.gov.uk/government/publications/water-and-treated-water/water-andtreated-water. 3) Source: https://www.umweltbundesamt.de/en/topics/water/drinking-water/small-scale-drinking-water-supplies#textpart-2 4) Source: Banzhaf et al. (2016) https://link.springer.com/article/10.1007/s13280-016-0848-8. * Extracted from http://nccph.netedit.info/docs/05_small_water_systems_ver_june2005.pdf, with newer data added where available (see notes). 5) Finland estimates that thousands of very small supplies exists but also do not know the number of users. Table A3.12 shows the number of sites that may use AFFFs in each country, totalling 694 airports and airbases, and 84,000 fire stations. Data are not complete, lacking small airports and possibly fire stations not intended primarily to serve the public and businesses generally, for example site emergency services at some industrial facilities, for 180 The cost of inaction linked to PFAS exposure example oil refineries. The omission of small airports may or may not be significant to the analysis: they may be less likely to use AFFFs (e.g. fire training may be coordinated at larger airports), but are also less likely to have effective containment in place. The aviation industry is also a significant user of PFAS in hydraulic fluid 407, though unlike AFFFs these are not deliberately released to the environment. Table A3.12: Number of public airports, military airbases and fire stations unrelated to aviation, by country Country Main airports (>150k passenger/y)1 Other airports (between 15k and 150k passenger/y)1 Military (air) bases2 Total Airports + Airbases Fire stations3 Total: EU,EFTA 318 137 239 694 84,099 Austria 6 0 5 11 5,199 Belgium 5 0 9 14 252 Bulgaria 3 1 8 12 220 Croatia 7 2 4 13 1,923 Cyprus 2 0 2 4 31 Czechia 3 2 7 12 7,561 Denmark 6 2 5 13 295 Estonia 1 1 1 3 187 Finland 9 8 7 24 988 France 44 18 38 100 6,897 Germany 25 16 25 66 33,460 Greece 20 14 13 47 275 Hungary 2 2 4 8 302 Iceland 1 0 1 2 Ireland 5 2 1 8 220 Italy 33 2 12 47 902 Latvia 1 0 1 2 92 Lithuania 3 0 3 83 Luxembourg 1 0 1 Malta 1 0 1 Netherlands 5 0 12 17 1,206 Norway 1 29 12 42 597 Poland 12 0 7 19 16,805 Portugal 8 5 9 22 473 Romaia 8 4 1 13 282 Slovakia 2 2 4 8 116 Slovenia 1 0 1 Spain 34 4 6 Sweden 19 11 5 Switzerland 18 1 6 2 1,359 44 35 1,002 25 1,319 UK 32 11 33 76 2,053 Note: 1) Eurostat Air Transport Statistics AIRP_TYP_ Number of Commercial Airports. 2) Sufficient official data has not been found, figures are based on Wikipedia for each country, and it is not clear which numbers include inactive bases. 3) International Association of Fire and Rescue Services (2017). World Fire Statistics. The omission of industrial facilities could be significant. RPA/BRE (2004) reported that 0.76 tonnes of PFOS based substance was held in Fire Authority inventories, whilst 23.7 tonnes was held in emergency stores at industrial complexes. Training at these sites 407 RPA and BRE Environment (2004). Perfluorooctane Sulphonate: Risk reduction strategy and analysis of advantages and drawbacks. Report no: J454/PFOS RRS. The cost of inaction linked to PFAS exposure 181 may or may not involve use of PFAS-containing materials, containment in training areas may or may not be effective. The number of landfill sites in each country is shown in Table A3.13, divided into landfills for hazardous waste, non-hazardous waste and inert waste. Hazardous waste covers materials that are toxic to humans, ecotoxic, carcinogenic, teratogenic, explosive etc (a complete list is provided in Annex III of the Directive on Hazardous Waste 408, and material meeting any of definitions is classified as hazardous). Hazardous waste landfills would be appropriate to any material significantly contaminated with PFAS, including filters and soils. Inert waste covers any material that will not react, degrade, dissolve or burn, with criteria set for leachability limits and hence should not include anything containing PFAS. Non-hazardous waste includes any other materials and would contain materials such as carpets, shoes, etc. that are contaminated with PFAS, but at lower concentration than material sent for hazardous waste disposal. Table A3.13: Number of landfill sites and incinerators in the EU28, Iceland and Norway Disposal - landfill (D1, D5, D12) Disposal - landfill for HW Disposal - landfill for non-HW Disposal - landfill for inert waste Disposal - incineration (D10) Austria 189 0 Belgium 64 9 Bulgaria 187 8 Croatia 146 0 Cyprus 7 1 Czech Republic 263 38 Denmark 41 5 Estonia 15 7 Finland 227 29 France 918 16 Germany 1,147 34 Greece 178 2 Hungary 111 13 Iceland Ireland 35 Italy 470 10 Latvia 13 2 Lithuania 14 0 Luxembourg 13 0 Malta 1 0 Netherlands 40 1 Norway 111 10 Poland 701 49 Portugal 60 2 Romania 129 7 Slovakia 118 11 Slovenia 37 1 Spain 520 35 Sweden 227 49 United Kingdom 594 26 Total 6,582 365 153 36 1 51 4 117 176 3 145 1 2 4 2 1 147 78 35 30 6 3 6 2 2 155 43 18 245 657 0 308 805 93 176 0 132 92 6 13 6 275 185 100 11 0 4 11 3 2 2 11 1 1 0 1 39 : 3 82 19 1 643 9 85 54 4 7 122 0 20 92 15 11 26 10 4 302 183 51 111 67 8 342 226 87 3,801 2,381 808 408 Council Directive of 12 December 1991 on hazardous waste (91 / 689 /EEC). 182 The cost of inaction linked to PFAS exposure Part 3: Full cost estimates by country Full cost estimates, by country, are provided in the following tables: the estimated costs for a basic screening programme (Table A3.14); the estimated costs of monitoring at contaminated sites (Table A3.15); estimated costs for water treatment works to reduce exposure to PFAS above possible limits (Table A3.16); estimated costs for soil remediation (Table A3.17); estimated costs for health assessment when contamination is found (Table A3.18); aggregated costs covering environmental screening, monitoring where contamination is found, water treatment, soil remediation and health assessment (Table A3.19). Table A3.14: Estimated costs for a basic screening programme N facilities for best estimate Best estimate, EUR million Low, EUR million Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total 806 148 39 141 25 736 78 131 184 2024 2771 89 119 7 413 300 69 151 21 19 111 179 2447 422 209 191 165 539 411 114 714 13,772 EUR 0.82 EUR 0.15 EUR 0.04 EUR 0.14 EUR 0.03 EUR 0.75 EUR 0.08 EUR 0.13 EUR 0.19 EUR 2.06 EUR 2.83 EUR 0.09 EUR 0.12 EUR 0.01 EUR 0.42 EUR 0.31 EUR 0.07 EUR 0.15 EUR 0.02 EUR 0.02 EUR 0.11 EUR 0.18 EUR 2.50 EUR 0.43 EUR 0.21 EUR 0.19 EUR 0.17 EUR 0.55 EUR 0.42 EUR 0.12 EUR 0.73 EUR 14.05 EUR 0.14 EUR 0.04 EUR 0.01 EUR 0.03 EUR 0.01 EUR 0.17 EUR 0.02 EUR 0.03 EUR 0.06 EUR 0.37 EUR 0.53 EUR 0.02 EUR 0.05 EUR 0.00 EUR 0.07 EUR 0.05 EUR 0.01 EUR 0.03 EUR 0.00 EUR 0.01 EUR 0.03 EUR 0.04 EUR 0.42 EUR 0.08 EUR 0.04 EUR 0.04 EUR 0.04 EUR 0.11 EUR 0.12 EUR 0.02 EUR 0.15 EUR 2.77 High, EUR million EUR 3.22 EUR 0.59 EUR 0.13 EUR 0.54 EUR 0.07 EUR 2.83 EUR 0.31 EUR 0.51 EUR 0.65 EUR 8.04 EUR 10.95 EUR 0.36 EUR 0.39 EUR 0.02 EUR 1.66 EUR 1.22 EUR 0.28 EUR 0.59 EUR 0.08 EUR 0.04 EUR 0.42 EUR 0.68 EUR 9.81 EUR 1.67 EUR 0.84 EUR 0.75 EUR 0.62 EUR 2.11 EUR 1.50 EUR 0.45 EUR 2.80 EUR 54.13 The cost of inaction linked to PFAS exposure 183 Best estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 3 samples, using best estimate of cost/sample for monitoring. 2. 5% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples. 3. Best estimate of costs adopted. Low estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 3 samples, using low cost/sample monitoring. 2. 1% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples. 3. Low estimate of costs adopted. High estimate assumptions 1. All airports and PFAS manufacturing sites are screened, assume 5 samples at each site, using high cost/sample for monitoring. 2. 10% of other facilities are screened (fire stations, waste water treatment works, large and small supplies, hazardous and MSW landfills), assume 3 samples. 3. High estimate of costs adopted. 184 The cost of inaction linked to PFAS exposure Table A3.15: Estimated costs of monitoring at contaminated sites N facilities for best estimate Best estimate, EUR millions, airfields and PFASmanufacturing only, EUR million Best estimate, all source categories included, EUR million Low, EUR millions High, EUR millions Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total 81 16 4 15 4 79 8 14 22 206 286 9 16 1 41 29 7 16 2 4 12 19 243 43 20 19 19 56 48 11 74 1,426 EUR 0.11 EUR 0.16 EUR 0.12 EUR 0.12 EUR 0.13 EUR 0.67 EUR 0.03 EUR 0.08 EUR 0.47 EUR 0.47 EUR 1.03 EUR 0.13 EUR 0.47 EUR 0.04 EUR 0.02 EUR 0.05 EUR 0.01 EUR 0.08 EUR 0.01 EUR 0.17 EUR 0.12 EUR 0.20 EUR 0.22 EUR 0.13 EUR 0.02 EUR 0.08 EUR 0.24 EUR 0.35 EUR 0.76 EUR 0.02 EUR 0.45 EUR 6.96 EUR 4.05 EUR 0.81 EUR 0.21 EUR 0.73 EUR 0.19 EUR 3.97 EUR 0.40 EUR 0.69 EUR 1.11 EUR 10.29 EUR 14.28 EUR 0.47 EUR 0.81 EUR 0.05 EUR 2.07 EUR 1.45 EUR 0.35 EUR 0.79 EUR 0.11 EUR 0.18 EUR 0.61 EUR 0.97 EUR 12.17 EUR 2.16 EUR 1.02 EUR 0.97 EUR 0.94 EUR 2.79 EUR 2.40 EUR 0.57 EUR 3.70 EUR 71.28 EUR 0.43 EUR 0.11 EUR 0.04 EUR 0.09 EUR 0.04 EUR 0.50 EUR 0.05 EUR 0.08 EUR 0.19 EUR 1.11 EUR 1.59 EUR 0.07 EUR 0.15 EUR 0.01 EUR 0.21 EUR 0.16 EUR 0.04 EUR 0.09 EUR 0.01 EUR 0.04 EUR 0.08 EUR 0.13 EUR 1.27 EUR 0.24 EUR 0.11 EUR 0.11 EUR 0.13 EUR 0.34 EUR 0.36 EUR 0.06 EUR 0.45 EUR 8.30 EUR 81.17 EUR 16.22 EUR 4.51 EUR 14.73 EUR 3.14 EUR 77.14 EUR 7.98 EUR 13.53 EUR 20.81 EUR 204.72 EUR 282.67 EUR 10.20 EUR 14.35 EUR 0.86 EUR 41.40 EUR 30.77 EUR 6.99 EUR 15.49 EUR 2.14 EUR 2.79 EUR 11.75 EUR 18.87 EUR 246.24 EUR 42.85 EUR 21.14 EUR 19.55 EUR 17.71 EUR 55.89 EUR 44.97 EUR 11.52 EUR 74.10 EUR 1,416 Best estimate assumptions 1. Assumed 20% of airports and PFAS manufacturing sites require monitoring pro- gramme, using best estimate cost/case for monitoring. 2. 0.5% of other facilities require monitoring. 3. Best estimate of costs adopted. Low estimate assumptions 1. Assumed 10% of airports and PFAS manufacturing sites require monitoring pro- gramme, using low estimate cost/case for monitoring. 2. 0.1% of other facilities require monitoring. 3. Low estimate of costs adopted. The cost of inaction linked to PFAS exposure 185 High estimate assumptions 1. Assumed 30% of airports and PFAS manufacturing sites require monitoring programme, using high estimate cost/case for monitoring. 2. 1% of other facilities require monitoring. 3. High estimate of costs adopted. Table A3.16: Estimated costs for water treatment works to reduce exposure to PFAS above possible limits Population affected, best estimate Best estimate, EUR millions Low, EUR millions High, EUR millions Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total 257,288 336,270 216,066 126,759 25,410 316,148 169,791 39,398 164,153 1,992,455 2,435,926 325,741 295,667 10,102 138,868 1,823,868 59,583 87,638 16,889 12,880 507,022 154,995 1,140,168 311,245 596,119 162,640 61,886 1,393,487 292,421 247,158 1,946,255 15,664,297 EUR 146 EUR 191 EUR 123 EUR 72 EUR 14 EUR 180 EUR 97 EUR 22 EUR 93 EUR 1,133 EUR 1,385 EUR 185 EUR 168 EUR 6 EUR 79 EUR 1,037 EUR 34 EUR 50 EUR 10 EUR 7 EUR 288 EUR 88 EUR 648 EUR 177 EUR 339 EUR 92 EUR 35 EUR 792 EUR 166 EUR 141 EUR 1,107 EUR 8,906 EUR 11 EUR 15 EUR 9 EUR 6 EUR 1 EUR 14 EUR 7 EUR 2 EUR 7 EUR 87 EUR 106 EUR 14 EUR 13 EUR 0 EUR 6 EUR 80 EUR 3 EUR 4 EUR 1 EUR 1 EUR 22 EUR 7 EUR 50 EUR 14 EUR 26 EUR 7 EUR 3 EUR 61 EUR 13 EUR 11 EUR 85 EUR 684 EUR 415 EUR 542 EUR 348 EUR 204 EUR 41 EUR 510 EUR 274 EUR 64 EUR 265 EUR 3,213 EUR 3,928 EUR 525 EUR 477 EUR 16 EUR 224 EUR 2,941 EUR 96 EUR 141 EUR 27 EUR 21 EUR 818 EUR 250 EUR 1,838 EUR 502 EUR 961 EUR 262 EUR 100 EUR 2,247 EUR 472 EUR 399 EUR 3,138 EUR 25,258 Best estimate assumptions 1. Assumed 3% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best estimate. 3. Assume best estimate cost per case for remediation. 4. Assume 4% discount rate on future maintenance costs. 186 The cost of inaction linked to PFAS exposure Low estimate assumptions 1. Assumed 1% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best es- timate. 3. Assume low estimate cost per case for remediation. 4. Assume 4% discount rate on future maintenance costs. High estimate assumptions 1. Assumed 5% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume 20 year maintenance programme for treatment works, based on best es- timate. 3. Assume high estimate cost per case for remediation. 4. Assume 4% discount rate on future maintenance costs. The cost of inaction linked to PFAS exposure 187 Table A3.17: Estimated costs for soil remediation N facilities for best estimate Best estimate, EUR millions, airfields and PFAS manufacturing only, EUR milions Best estimate, all source categories included, EUR milions Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total 81 16 4 15 4 79 8 14 22 206 286 9 16 1 41 29 7 16 2 4 12 19 243 43 20 19 19 56 48 11 74 1,426 11.0 16.0 12.0 12.0 13.0 67.0 3.0 8.0 47.0 47.0 103.0 13.0 47.0 4.0 2.0 5.0 1.0 8.0 1.0 17.0 12.0 20.0 22.0 13.0 2.0 8.0 24.0 35.0 76.0 2.0 45.0 696 EUR 405 EUR 81 EUR 21 EUR 73 EUR 19 EUR 397 EUR 40 EUR 69 EUR 111 EUR 1,029 EUR 1,428 EUR 47 EUR 81 EUR 5 EUR 207 EUR 145 EUR 35 EUR 79 EUR 11 EUR 18 EUR 61 EUR 97 EUR 1,217 EUR 216 EUR 102 EUR 97 EUR 94 EUR 279 EUR 240 EUR 57 EUR 370 EUR7,128 Low, EUR millions EUR 5.1 EUR 1.3 EUR 0.5 EUR 1.1 EUR 0.5 EUR 6.0 EUR 0.5 EUR 1.0 EUR 2.2 EUR 13.3 EUR 19.1 EUR 0.9 EUR 1.8 EUR 0.1 EUR 2.5 EUR 2.0 EUR 0.4 EUR 1.1 EUR 0.1 EUR 0.5 EUR 1.0 EUR 1.6 EUR 15.2 EUR 2.8 EUR 1.3 EUR 1.3 EUR 1.6 EUR 4.1 EUR 4.3 EUR 0.7 EUR 5.4 EUR 100 High, EUR millions EUR 8,117 EUR 1,622 EUR 451 EUR 1,473 EUR 314 EUR 7,714 EUR 798 EUR 1,353 EUR 2,081 EUR 20,472 EUR 28,267 EUR 1,020 EUR 1,435 EUR 86 EUR 4,140 EUR 3,077 EUR 699 EUR 1,549 EUR 214 EUR 279 EUR 1,175 EUR 1,887 EUR 24,624 EUR 4,285 EUR 2,114 EUR 1,955 EUR 1,771 EUR 5,589 EUR 4,497 EUR 1,152 EUR 7,410 EUR 141,613 Best estimate assumptions 1. Assumed 20% of airports and PFAS manufacturing sites require remediation. 2. 0.5% of other facilities require remediation. 3. Best estimate of costs adopted. Low estimate assumptions 1. Assumed 10% of airports and PFAS manufacturing sites require remediation. 2. 0.1% of other facilities require remediation. 3. Low estimate of costs adopted. 188 The cost of inaction linked to PFAS exposure High estimate assumptions 4. Assumed 30% of airports and PFAS manufacturing sites require remediation. 5. 1% of other facilities require remediation. 6. High estimate of costs adopted. Table A3.18: Estimated costs for health assessment when contamination is found Population affected, best estimate Best estimate, EUR millions Low, EUR millions High, EUR millions Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total 257,288 336,270 216,066 126,759 25,410 316,148 169,791 39,398 164,153 1,992,455 2,435,926 325,741 295,667 10,102 138,868 1,823,868 59,583 87,638 16,889 12,880 507,022 154,995 1,140,168 311,245 596,119 162,640 61,886 1,393,487 292,421 247,158 1,946,255 15,664,297 EUR 13 EUR 17 EUR 11 EUR 6 EUR 1 EUR 16 EUR 8 EUR 2 EUR 8 EUR 100 EUR 122 EUR 16 EUR 15 EUR 1 EUR 7 EUR 91 EUR 3 EUR 4 EUR 1 EUR 1 EUR 25 EUR 8 EUR 57 EUR 16 EUR 30 EUR 8 EUR 3 EUR 70 EUR 15 EUR 12 EUR 97 EUR 783 EUR 0.43 EUR 0.56 EUR 0.36 EUR 0.21 EUR 0.04 EUR 0.53 EUR 0.28 EUR 0.07 EUR 0.27 EUR 3.32 EUR 4.06 EUR 0.54 EUR 0.49 EUR 0.02 EUR 0.23 EUR 3.04 EUR 0.10 EUR 0.15 EUR 0.03 EUR 0.02 EUR 0.85 EUR 0.26 EUR 1.90 EUR 0.52 EUR 0.99 EUR 0.27 EUR 0.10 EUR 2.32 EUR 0.49 EUR 0.41 EUR 3.24 EUR 26 EUR 41 EUR 53 EUR 34 EUR 20 EUR 4 EUR 50 EUR 27 EUR 6 EUR 26 EUR 315 EUR 386 EUR 52 EUR 47 EUR 2 EUR 22 EUR 289 EUR 9 EUR 14 EUR 3 EUR 2 EUR 80 EUR 25 EUR 181 EUR 49 EUR 94 EUR 26 EUR 10 EUR 221 EUR 46 EUR 39 EUR 308 EUR 2,480 The cost of inaction linked to PFAS exposure 189 Table A3.19: Aggregated costs covering environmental screening, monitoring where contamination is found, water treatment, soil remediation and health assessment. Best estimate, EUR millions Low, EUR millions High, EUR millions Austria Belgium Bulgaria Croatia Cyprus Czechia Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Slovakia Slovenia Spain Sweden Switzerland UK Total EUR 569 EUR 290 EUR 155 EUR 152 EUR 35 EUR 597 EUR 145 EUR 94 EUR 214 EUR 2,274 EUR 2,952 EUR 249 EUR 264 EUR 12 EUR 295 EUR 1,275 EUR 72 EUR 134 EUR 21 EUR 26 EUR 375 EUR 194 EUR 1,937 EUR 411 EUR 472 EUR 199 EUR 133 EUR 1,144 EUR 423 EUR 210 EUR 1,579 EUR 16,902 EUR 17 EUR 16 EUR 10 EUR 7 EUR 2 EUR 21 EUR 8 EUR 3 EUR 10 EUR 105 EUR 132 EUR 16 EUR 15 EUR 1 EUR 9 EUR 85 EUR 3 EUR 5 EUR 1 EUR 1 EUR 24 EUR 9 EUR 69 EUR 17 EUR 28 EUR 9 EUR 5 EUR 68 EUR 18 EUR 12 EUR 94 EUR 821 EUR 8,656 EUR 2,234 EUR 838 EUR 1,712 EUR 362 EUR 1,106 EUR 1,437 EUR 2,393 EUR 24,213 EUR 32,874 EUR 1,607 EUR 1,973 EUR 105 EUR 4,429 EUR 6,339 EUR 811 EUR 1,720 EUR 246 EUR 305 EUR 2,085 EUR 2,181 EUR 26,899 EUR 4,880 EUR 3,191 EUR 2,263 EUR 1,899 EUR 8,115 EUR 5,061 EUR 1,601 EUR 10,933 EUR 170,821 Best estimate assumptions 1. Assumed 3% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume best estimate cost per case for remediation. Low estimate assumptions 1. Assumed 1% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume low estimate cost per case for remediation. 190 The cost of inaction linked to PFAS exposure High estimate assumptions 1. Assumed 5% of the population are exposed to excess levels of PFAS via drinking water. 2. Assume high estimate cost per case for remediation. The cost of inaction linked to PFAS exposure 191 Nordic Council of Ministers Nordens Hus Ved Stranden 18 DK-1061 Copenhagen K www.norden.org THE COST OF INACTION PFAS (per and polyfluoroalkylsubstances) are known to be extremely difficult to degrade in the environment and to be bioaccumulative and toxic. Exposure to PFAS is suspected to increase the risk of adverse health effects, such as impacts on the thyroid gland, the liver, fat metabolism and the immune system. This study estimates the socioeconomic costs that may result from impacts on human health and the environment from the use of PFAS. Better awareness of the costs and problems associated with PFAS exposure will assist decision-makers and the general public to make more efficient and timely risk management decisions. Findings indicate that the costs are substantial, with annual health-related costs estimated to 2.8 - 4.6 billion EUR for the Nordic countries and 52 - 84 billion EUR for all EEA countries. Overall non-health costs are estimated at 46 million - 11 billion EUR for the Nordic countries. Upon request the excel spreadsheets used for the monetarisation and valuation in this report can also be provided along with a guidance on how to use the estimation of costs for value transfer. Please contact any of the consultants or members of the steering group from the Swedish Chemicals Agency or the Danish Environmental Protection Agency if you are interested in receiving these excel spreadsheets.