PFAS Regulation of the EU

Overview and basic data processing

Per- and polyfluoroalkyl substances (PFAS) are an important class of synthetic chemicals with many industrial and non-industrial applications. Due to environmental concerns, the EU represented by the European Chemicals Agency (ECHA) submitted a universal PFAS restriction proposal. Following the initial proposal a six-month consultation phase of interested parties lead to the submission of more than 5600 comments from the public by the end on 25 September 2023. These comments, as far as not confidential, were made available in 124 .docx-files (MS Word) on 2 November 2023.^[1] The content presented here is published with the intent of making the data more easily accessible.^[2]

Unfortunately, the information in the files is not easily accessible, especially since the only filtering can be performed in the separate index file (Excel-file), in which no information on industry segments of the submitting companies is provided. Furthermore, many additional documents are embedded into the separate .docx-files, a topic not addressed by this document.

However, in consequence of the way the results from the consultation phase are published, information retrieval for industrial players is unnecessarily complicated resulting in an even greater information gap between ECHA and submitting organisations on the overall process. The intention here is to reduce this information gap by providing means to generate a comprehensive searchable table with all non-confidential content not submitted in attached files. Furthermore, from the result it is easily possible to generate tables for a relational database to link it to additional information.

The information requests by the ECHA for this consultation phase, as provided at the beginning of each .docx-file, were:

Sectors and (sub-)uses: Please specify the sectors and (sub-)uses to which your comment applies according to the sectors and (sub-)uses identified in the Annex XV restriction report (Table 9). If your comment applies to several sectors and (sub-)uses, please make sure to specify all of them.
Emissions in the end-of-life phase: The environmental impact assessment does not cover emissions resulting from the end-of-life phase. To get a better understanding of the extent of the resulting underestimation, (sub-)use-specific information is requested on emissions across the different stages of the lifecycle of products, i.e. the manufacture phase, the use phase and the end-of-life phase. Please provide justifications for the representativeness of the provided information. In particular:
1. Please provide, at the (sub-)use level, an indication of the share of emissions (as percentages) attributable to these three different stages. An indication of annual emission volumes in the end-of-life phase at sector or sub-sector level would also be appreciated.
2. If possible, please provide for each (sub-)use what share of the waste (as percentages) is treated through incineration, landfilling and recycling. Please provide information to justify the estimates as well as information on the form of recycling referred to.
Emissions in the end-of-life phase: With respect to waste management options, additional information is requested on the effectiveness of incineration under normal operational conditions (for different waste types, e.g. hazardous, municipal) with respect to the destruction of PFAS and the prevention of PFAS emissions.
Impacts on the recycling industry: To get an understanding of the impacts of the proposed restriction on the recycling industry, information is requested on:
1. The impacts that the concentration limits proposed in paragraph 2 of the proposed restriction entry text (see table starting on page 4 of the summary of the Annex XV restriction report) have on the technical and economic feasibility of recycling processes (together with a clear indication on the waste streams to which the described impacts relate).
2. The measures that recyclers would need to take to achieve the proposed concentration limits.
3. The costs associated with these measures.
Proposed derogations – Tonnage and emissions: Paragraphs 5 and 6 of the proposed restriction entry text (see table starting on page 4 of the summary of the Annex XV restriction report) include several proposed derogations. For these proposed derogations, information is requested on the tonnage of PFAS used per year and the resulting emissions to the environment for the relevant use. Please provide justifications for the representativeness of the provided information.
Missing uses – Analysis of alternatives and socio-economic analysis: Several PFAS uses have not been covered in detail in the Annex XV restriction report (see uses highlighted in blue and orange in Table A.1 of Annex A of the Annex XV restriction report). In addition, some relevant uses may not have been identified yet. For such uses, specific information is requested on alternatives and socio-economic impacts, covering the following elements:
1. The annual tonnage and emissions (at sub-sector level) and type of PFAS associated with the relevant use.
2. The key functionalities provided by PFAS for the relevant use.
3. The number of companies in the sector estimated to be affected by the restriction.
4. The availability, technical and economic feasibility, hazards and risks of alternatives for the relevant use, including information on the extent (in terms of market shares) to which alternative-based products are already offered on the EU market and whether any shortages in the supply of relevant alternatives are expected.
5. For cases in which alternatives are not yet available, information on the status of R&D processes for finding suitable alternatives, including the extent of R&D initiatives in terms of time and/or financial investments, the likelihood of successful completion, the time expected to be required for substitution (including any relevant certification or regulatory approvals) and the major challenges encountered with alternatives which were considered but subsequently disregarded.
6. For cases in which substitution is technically and economically feasible but more time is required to substitute:
  1. the type and magnitude of costs (at company level and, if available, at sector level) associated with substitution (e.g. costs for new equipment or changes in operating costs);
  2. the time required for completing the substitution process (including any relevant certification or regulatory approvals);
  3. information on possible differences in functionality and the consequences for downstream users and consumers (e.g. estimations of expected early replacement needs or expected additional energy consumption);
  4. information on the benefits for alternative providers.
7. For cases in which substitution is not technically or economically feasible, information on what the socio-economic impacts would be for companies, consumers, and other affected actors. If available, please provide the annual value of EU sales and profits of the relevant sector, and employment numbers for the sector.
Potential derogations marked for reconsideration – Analysis of alternatives and socio-economic analysis: Paragraphs 5 and 6 of the proposed restriction entry text (see table starting on page 4 of the summary of the Annex XV restriction report) include several potential derogations for reconsideration after the consultation (in [square brackets]). These are uses of PFAS where the evidence underlying the assessment of the substitution potential was weak. The substitution potential is determined on the basis of i) whether technically and economically feasible alternatives have already been identified or alternative-based products are available on the market at the assumed entry into force of the proposed restriction, ii) whether known alternatives can be implemented before the transition period ends (taking into account time requirements for substitution and certification or regulatory approval), and iii) whether known alternatives are available in sufficient quantities on the market at the assumed entry into force to allow affected companies to substitute.
A summary of the available evidence as well as the key aspects based on which a derogation is potentially warranted are presented in Table 8 in the Annex XV restriction report, with further details being provided in the respective sections in Annex E.
To strengthen the justifications for a derogation for these uses, additional specific information is requested on alternatives and socio-economic impacts covering the elements described in points a) to g) in question 6 above.
Other identified uses – Analysis of alternatives and socio-economic analysis: Table 8 in the Annex XV restriction report provides a summary of the identified sectors and (sub-)uses of PFAS, their alternatives and the costs expected from a ban of PFAS. More details on the available evidence are provided in the respective sections in Annex E.
For many of the (sub-)uses, the information on alternatives and socio-economic impacts was generic and mainly qualitative. In particular, evidence on alternatives was inconclusive for some applications falling under the following (sub-)uses: technical textiles, electronics, the energy sector, PTFE thread sealing tape, non-polymeric PFAS processing aids for production of acrylic foam tape, window film manufacturing, and lubricants not used under harsh conditions.
More information is needed on alternatives and socio-economic impacts to conclude on substitution potential, proportionality, and the need for specific time-limited derogations. Therefore, specific information (if not already included in the Annex XV restriction report or covered in the questions above) is requested on alternatives and socio-economic impacts covering the elements listed in points a) to g) in question 6 above.
Degradation potential of specific PFAS sub-groups: A few specific PFAS sub-groups are excluded from the scope of the restriction proposal because of a combination of key structural elements for which it can be expected that they will ultimately mineralize in the environment. RAC would appreciate to receive any further information that may be available regarding the potential degradation pathways, kinetics or produced metabolites in relevant environmental conditions and compartments for trifluoromethoxy, trifluoromethylamino- and difluoromethanedioxy-derivatives.
Analytical methods: Annex E of the Annex XV restriction report contains an assessment of the availability of analytical methods for PFAS. Analytical methods are rapidly evolving. Please provide any new or additional information on new developments in analytics not yet considered in the Annex XV restriction report.

Let’s assume, that all .docx-files (approximately 2.5 GB) are already downloaded into a folder called “ECHA_Word-files”. The first task is to extract all data contained in the tables and merge them into a single table prior to complete data cleaning. Subsequently, the resulting table is processed to clean up the messy format and prepare it for further processing and data analysis. Below you find the necessary R code to perform these tasks. It is certainly not the most elegant solution, but it works.

library(dplyr)
library(tidyr)
library(docxtractr)
library(stringr)

# Load required libraries

docx_files <- list.files(path = "./ECHA_Word-files",
                         pattern = "\\.docx$",
                         full.names = T)

# Initialize an empty list to store the tables
table_list <- list()
# Loop through each docx file
for (file in docx_files) {
# Read the docx file
doc <- read_docx(file)

# Extract all tables from the docx file
tables <- docx_extract_all_tbls(doc, guess_header = FALSE, preserve = TRUE)

# Add the tables to the list
table_list <- c(table_list, tables)
}

# Combine all tables into one data frame
combined_table <- do.call(rbind, table_list) %>%
  slice(-1) %>%
  rename(ID = V1, SubmissionData = V2, Comments = V3) %>%
  mutate_all(~replace(., . == "", NA)) %>%
  fill(ID, .direction = "down")

# Split table to process column 2 and 3 separately

table_submission_data <- combined_table %>%
  select(ID, SubmissionData) %>%
  drop_na() %>%
  separate_rows(SubmissionData, sep = "\n") %>%
  mutate(SubmissionData = str_trim(SubmissionData),
         Title = case_when(SubmissionData == "Date:" ~ "Date",
                           SubmissionData == "Content:" ~ "Content",
                           SubmissionData == "Type:" ~ "Type",
                           SubmissionData == "Org. type:" ~ "OrganizationType",
                           SubmissionData == "Org. name:" ~ "OrganizationName",
                           SubmissionData == "Org. country:" ~ "OrganizationCountry",
                           SubmissionData == "Country:" ~ "SubmitterCountry",
                           SubmissionData == "Attachment:" ~ "Attachment",
                           SubmissionData == "Privacy statement:" ~ "PrivacyStatement",
                           SubmissionData == "Company name confidential:" ~ "ConfidentialCompanyName",
                           TRUE ~ NA
                           )
         ) %>%
  fill(Title, .direction = "down") %>%
  filter(!SubmissionData %in% c("", "Date:", "Content:", "Type:", "Org. type:", "Org. name:", "Org. country:", "Country:", "Attachment:", "Privacy statement:", 
                                "Company name confidential:", "\\s")) %>%
  mutate(temp = if_else(Title == "Content", SubmissionData, NA)) %>%
  group_by(ID) %>%
  mutate(ContentTypes = paste(temp[!is.na(temp)], collapse = "; ")) %>%
  filter(Title != "Content") %>%
  pivot_wider(names_from = Title, values_from = SubmissionData) %>%
  ungroup() %>%
  select(-temp)
  
table_submission_comments <- combined_table %>%
  select(ID, Comments) %>%
  drop_na() %>%
  mutate(Title = case_when(str_detect(Comments, "^General Comments:\n") ~ "GeneralComments",
                           str_detect(Comments, "^Answer to specific info request 1:\n") ~ "AnswerQ1",
                           str_detect(Comments, "^Answer to specific info request 2:\n") ~ "AnswerQ2",
                           str_detect(Comments, "^Answer to specific info request 3:\n") ~ "AnswerQ3",
                           str_detect(Comments, "^Answer to specific info request 4:\n") ~ "AnswerQ4",
                           str_detect(Comments, "^Answer to specific info request 5:\n") ~ "AnswerQ5",
                           str_detect(Comments, "^Answer to specific info request 6:\n") ~ "AnswerQ6",
                           str_detect(Comments, "^Answer to specific info request 7:\n") ~ "AnswerQ7",
                           str_detect(Comments, "^Answer to specific info request 8:\n") ~ "AnswerQ8",
                           str_detect(Comments, "^Answer to specific info request 9:\n") ~ "AnswerQ9",
                           str_detect(Comments, "^Answer to specific info request 10:\n") ~ "AnswerQ10",
                           TRUE ~ NA
                           ),  
           Comments = case_when(Title == "GeneralComments" ~ str_sub(Comments, start = 18),
                                Title %in% c("AnswerQ1", "AnswerQ2", "AnswerQ3", "AnswerQ4", "AnswerQ5", "AnswerQ6", "AnswerQ7", "AnswerQ8", "AnswerQ9") ~ str_sub(Comments, start = 36),
                                Title == "AnswerQ10" ~ str_sub(Comments, start = 37),
                                TRUE ~ Comments
                                )
         ) %>%
  group_by(ID) %>%
  pivot_wider(names_from = Title, values_from = Comments) %>%
  ungroup()

combined_table_cleaned <- table_submission_data %>%
  left_join(table_submission_comments, by = "ID") %>%
  select(ID, Date, Type, OrganizationType, OrganizationName, ConfidentialCompanyName, OrganizationCountry, SubmitterCountry, ContentTypes, Attachment, PrivacyStatement, GeneralComments,
         AnswerQ1, AnswerQ2, AnswerQ3, AnswerQ4, AnswerQ5, AnswerQ6, AnswerQ7, AnswerQ8, AnswerQ9, AnswerQ10) %>%
  mutate(ConfidentialCompanyName = if_else(is.na(ConfidentialCompanyName), "no", tolower(ConfidentialCompanyName)))

Now we have desired table which can be used for further analysis. In a database additionally the documents embedded the .docx-files could be stored, making them accessible as well.

Basic data analysis

All subsequent analysis is based on this dataset without any further cleaning. Overall, 5642 submissions were filed originating from member states of the EU, various organizations and individuals.

EU PFAS-Regulation submitting entities — Figure based on data from ECHA processed in the way described above.

All these organisations and individuals expressed their opinion or provided content of vastly different from experimental results provided in attached files to just a comment “OK”. The ECHA assigned content types to all submissions with any content of noticeable value.

EU PFAS-Regulation content types — Figure based on data from ECHA processed in the way described above.

When looking at the origin of individuals, an interesting pattern emerges:

Origin of individuals submitting on EU PFAS-Regulation and the content provided — Figure based on data from ECHA processed in the way described above.

The vast majority of all submissions by individuals originate from Sweden, indicating that the level of concern highly depends on public discussions within individual countries. Furthermore, people were often rather sloppy in entering the content into the appropriate fields of the submission; some individual even made entries with “Lorem ipsum…”. 142 entries did contain neither a comment nor an answer to any of the questions, mostly because the general comment was falsely entered in the field for the privacy statement. Furthermore, 1271 submissions only contained a comment without answering any of the questions. The number of answers provided to the individual questions from the remaining 129 submissions by individuals are shown above. Overall, individual submissions are mostly in favor of a stricter regulation of PFAS usage. The content provided by individuals is mainly of low quality with little to no evidence provided regarding the claimed PFAS effects (no references to respective studies etc.). The high similarity between many submissions from Sweden further suggests an organized campaign with some template or reference text provided or multiple submissions by some individuals.

When looking into materials in focus (without considering redacted entries and attached files), the following frequency occurs:

Materials and substances mentioned in content provided in submissions (without attachments) — Figure based on data from ECHA processed in the way described above.

Looking at the numbers, particularly fluoropolymers (1424, 25.24%) are mentioned in many submissions by the industry, in particular PTFE (1390, 24.64%), PVDF (491, 8.7%) and fluoroelastomers (775, 13.74%). Notice that for some applications also potential alternatives like polyether ether ketone (PEEK) or polybutylene terephthalate (PBT) are mentioned in the submissions. The data only provide an approximate number, as the non-English contributions have been neither translated nor were all the different nomenclatures or other spellings considered, and, in addition to that, were all the additional files submitted negleted.

Footnotes

Available at https://echa.europa.eu/de/comments-submitted-to-date-on-restriction-report-on-pfas.
I decided not to make the resulting table publicly available due to potential legal issues. If you are interested in the table, please contact me via LinkedIn https://www.linkedin.com/in/norman-hendrik-riedel-2550531a or using the email on the contact page.