Changelog

11 September 2026 (A full accuracy sweep, and three things we got wrong)

We reviewed every page on this site against the evidence we now hold. Several things needed correcting, and three of them were our own errors rather than simply out-of-date information. They are set out first, because that is the order they deserve.

We said we had put questions to bodies we had not yet contacted. Sixteen posts carried wording such as “we are asking this openly, and putting it directly to WSP and the Environment Agency”, or to Lancashire County Council and the Health Cell. Our own internal tracker recorded those requests as not sent, and a check of our outbound correspondence confirmed it: they had not been sent. The wording now says what is true, that we are preparing to put these questions formally to the bodies named. Nothing we published about any organisation’s conduct depended on this, and no such claim has been changed. But it mattered, because a body that has never been approached can say so, and it would be right. We had written the rule for ourselves before we broke it: this site should not imply an organisation is ducking a question we have not actually asked.

A soil benchmark we quoted does not exist. Several pages compared PFOA soil readings against “the Dutch screening value of 3.8 µg/kg”. We can find that figure in no RIVM publication. It appears to have been introduced in error rather than mis-transcribed from a real document. We caught this on 2 August and removed it from The Numbers, but we did not check the rest of the site, and it remained on five other pages, including in one post’s list of sources, where a figure that does not exist was presented as a citation. Every instance is now replaced with the screening values WSP actually applies, from its own Phase 2 report for the Environment Agency: 1.4 µg/kg for allotment soil, and 7.6 µg/kg for residential gardens where vegetables are grown. The soil readings themselves are unchanged; only the benchmark they are measured against has been corrected. Where we previously said all 25 samples at one property exceeded the benchmark, we now state the position against both real values, and note that the lowest reading of the 25 sits just below the residential figure.

Our timeline described itself inaccurately. It stated that three items previously listed as forthcoming, the Phase 3 reports, the HSE consultation outcome and the Leigh Day town hall, “have now passed and are recorded above where they occurred”. They are not recorded. The same page still described all three in the future tense. The timeline now says plainly that its dated entries run to March 2026, that those three items passed their dates without our establishing what happened, and that we would rather say so than imply a record we do not hold. A short summary of developments since March has been added, with each item linking to the investigation that carries its sources.

A correction from 3 August that we never recorded here. In early August we corrected a false claim that the Lancashire County Council Health Cell kidney cancer report was unpublished and reached only through legal correspondence. It is published in full by Wyre Council. We changed three posts at the time but did not record it on this page, which our own standards require. We are recording it now rather than back-dating it.

The rest of the sweep brought pages up to date with evidence we already held.

The homepage was still advertising an information event held on 21 May, and the What To Do page was still listing drop-in sessions from March as the next ones, with an invitation to turn up. Both now carry the current position, and say plainly that no further sessions have been announced.

Several pages described AGC Chemicals Europe as the current operator with no mention of the closure of production announced on 13 July 2026. They now note the announcement, that no closure date has been made public, and Wyre Council’s statement that the investigation is unaffected.

Sandringham Avenue Allotments were variously described as “to close” and “closed pending results”, in two places on the same page. Wyre Council’s letter to residents of 12 August 2026 states that the Environment Agency has concluded that this site, like Occupation Road, meets the criteria to be determined as contaminated land under Part 2A. Every page now says so, and the guidance-change table has a row for that letter. The same letter records that elevated PFAS was found in every egg sample taken from the four poultry keepers tested.

The health page said in four places that no study had examined local cancer rates in relation to the site, and that age-adjusted small-area data had not been published. That is no longer true for kidney cancer: the Health Cell assessment published figures for the neighbourhood immediately south-east of the site, 14 cases against 6 expected, a standardised incidence ratio of 218. We have narrowed those statements rather than deleted them, because they still hold for every other cancer and for the non-cancer conditions associated with PFAS, and the page now says which is which.

The Numbers gave PFOA in surface water as “12,100 to 20,624 ng/L”. That range spliced the single value from one study to the maximum from another and appeared in neither. It now reads “up to 20,624 ng/L”, with the two studies given separately.

One post dated the Health Cell report to 16 June 2026, which is when it was forwarded to Leigh Day, rather than 12 May 2026, its own review date. Every other post had this right.

Later the same day we put the queued articles on the October 2024 fire through three independent reviews. Two of their findings landed on pages already published, and we have corrected both.

Our timeline said the fire “had not been reported at the time”. The regulators were told: the Environment Agency was notified that evening, and AGC filed its statutory notification the following day. What had not happened was any public report. The page now says so, and gives the notification dates. That error was introduced earlier the same day, in the correction described above.

Our Open Questions page said of AGC’s statement that no enforcement action followed, “We have no reason to doubt it”. That was written before the Environment Agency’s record was disclosed. The record shows a breach of a permit condition, categorised C3, with a warning issued and a required action due 25 March 2025. Whether a warning amounts to “enforcement action” depends on what the term covers, and the Agency’s own form describes enforcement action as something that can include a formal caution, prosecution, the service of a notice, or suspension or revocation of a permit. We now set out the record and both readings rather than offering a view.

Two quotations that we had flagged internally as unverified turned out to be sound. In both cases we had saved the documents attached to a disclosure but not the covering letter that carried the authority’s actual answer — including, in one case, the answer to the question the whole request was built around. Both letters are now archived. We mention it because the failure is an instructive one: for six weeks our own records suggested we could not support two things we had published, and the proof was sitting on a public page we had linked to all along.

One post carried two bracketed notes-to-self that should never have reached publication, one asking for a source for the 71 properties figure, the other for the planning moratorium correspondence. Both are now sourced. On the second, we had the document all along, in our own archive and on our own timeline. We have also corrected how we described it: saying the government “refused” the request overstates it, because the reply pointed the council to planning powers it already holds rather than blocking it from acting, and the shorter version wrongly frames central government as the obstacle.

3 September 2026 (Groundwater: sampling is happening, results are not published)

We have expanded what “The Numbers” says about groundwater.

The page said there was no published data on PFAS in groundwater beneath or adjacent to the estate. That was accurate and still is. But it was incomplete, because a reader could reasonably have taken it to mean that nobody is testing groundwater at all. That is not the case.

AGC Chemicals Europe states on its own website that it has “commissioned a specialist third party to conduct the voluntary on-site investigation”, which it describes as an intrusive investigation within the boundaries of land it owns, “to sample and test soil, groundwater and surface water”. It says Stage 1 is complete, that Stage 2 began in June 2026, and that the details of each stage are agreed with Wyre Council and the Environment Agency.

So groundwater is being sampled. What is missing is not the testing. It is the results, which have not been published, and the identity of the third party, which is not named.

We have corrected our own internal research notes, which had gone further than the site did and said groundwater “has never been assessed”. That was wrong, and we have said so in our corrections log.

Source: AGC Chemicals Europe, Site Investigation page, accessed 3 September 2026.

Citations

A link check of every external citation on the site found three dead links. All three have been dealt with. The corrections were made on 7 August and are published today.

  • Removed. The “Sources and Evidence” page listed a Guardian article at /environment/2026/feb/03/pfas-action-plan. That page does not exist and, as far as we can establish, never did. The link has been removed along with the heading that held it. No claim on the site rested on it: the UK PFAS plan it referred to is still cited on the same page from its GOV.UK source. We do not know how a citation to a non-existent page came to be on the site, and we are not going to guess. It should not have been there.
  • Repointed. “The Numbers” page attributed the pre-publication figure of over 400,000 ng/L to a Guardian article dated 17 February 2023. The quote is real and the reporting is real, but the date and the address were both wrong. It is the article of 24 February 2023, now linked directly. The figure itself was already marked on the page as pre-peer-review, and we still cite the published figure of about 40,000 ng/L instead.
  • Repointed. The Food Standards Agency link pointed at cot.food.gov.uk, which now returns “410 Gone”. This is not a fault at our end or theirs: the Committee on Toxicity’s old website has been retired in full. The citation now goes to the preserved copy in the UK Government Web Archive, with the Committee’s current GOV.UK page alongside it, and the page says plainly that the source is an archived one.

Contact form

The contact form has been rebuilt. It now shows a live character count as you type, saves a draft in your own browser so a closed tab or a failed send does not lose what you wrote, reports any failure on the page instead of sending you to an error screen, and records how long your message was so that we can tell if anything arrives short.

There is no length limit. Long accounts are welcome and are the most useful thing we receive.

If you sent us a long message before 12 August 2026 and had no reply, please send it again. At least one detailed account reached us cut off part way through a sentence, and we could not recover the rest of it. We are sorry. The form should not have been capable of that.

2 August 2026 (Major correction pass: source verification across the whole site)

We ran a full verification sweep of every page against the primary sources they cite, fetching and reading the sources rather than relying on our own earlier summaries of them. It found errors we should have caught before publishing. They are set out here in full because a site that asks a regulator and a company to show their working has to show its own.

The main correction

We had placed the peak PFOA reading of 20,624 ng/L at discharge point W1, the main permitted outfall. That is wrong. Dalmijn et al. (2025) record W1 as the lowest of every point they sampled, at 30 and 33 ng/L total PFAS. The high readings, ranging from 891 to 20,624 ng/L, were at three points north of the plant: a discharge beside a lagoon, a land drain, and a leaking section of discharge pipe. The paper identifies that lagoon as an effluent pit used by ICI, which operated the site until 1999, and records that it could not be accessed during sampling.

The error appeared in five places across three pages and has been corrected in all of them. We had also attributed the 1,744 ng/L EEA-NH4 figure to discharge point W2; the paper reports it for Hillylaid Pool water, and gives no separate figure for the discharge point.

We had additionally not reported what the paper says about cause. It attributes the high PFOA to use at the plant between 1955 and 2012 and to a reservoir remaining within the site, with re-emission as the mechanism, rather than to current discharge. That is now stated wherever the figures appear. AGC’s own reporting to the US EPA shows it was emitting PFOA as operator until 2012, and the published record does not break the total down between the two operators.

Citations

Every one of the six PubMed links on the “What is PFAS” page supporting the C8 Science Panel’s findings pointed to an unrelated paper. All six have been replaced with citations checked against the PubMed record. Kidney cancer was also credited to the wrong study. “Pre-eclampsia” has been corrected to “pregnancy-induced hypertension”, which is the Panel’s own term.

Figures

  • Total organofluorine “around 400,000 ng/L” corrected to total PFAS of about 40,000 ng/L. The cited paper did not measure organofluorine. The higher figure came from a pre-publication quote.
  • The EFSA limit was given as 4.4 µg/kg. It is 4.4 ng per kg of bodyweight per week.
  • A dioxins figure was given in micrograms; the source states nanograms.
  • A thermal oxidiser “offline 43 days” conflated 56 days of outages with 43 invalid monitoring days.
  • Several dates were wrong, including the publication date of the Dalmijn paper, the date of a company name change, the date of Phase 2 soil fieldwork, and the date of a chemical spill.

Attribution

We had described Professor Crispin Halsall of Lancaster University, an independent academic with no connection to the company, as an “AGC spokesperson”. We had presented a company’s statement to journalists as data submitted to a regulator. We had attributed to ECHA classifications that were the Environment Agency’s assessment and a supplier’s self-classification. We had stitched two clauses of a Guardian article into a single quotation. All corrected.

Claims we have withdrawn

  • That the soil report’s PDF carried the same name in two independent metadata fields. It appears in one field, echoed by the software into a second block. The rest of that post stands.
  • That March 2026 was the first time the contamination reached the Parliamentary record. It was raised in June 2025.
  • Language in our coverage of the 2000 industry meeting that asserted the company “chose” to continue, and an anonymous nickname aimed at an individual. We report what the minutes record and what happened afterwards, and we do not assert intent.

A full internal record of every change, with the verbatim source text relied on, is kept alongside the site.

23 July 2026 (Precision correction: fire duration wording)

Correction

  • Fire article (agc-fire-three-month-shutdown.md) and Questions tracker: In two places the article used the shorthand “a three-month fire.” This was imprecise and could be misread: the fire itself was short-lived; what lasted three months was the shutdown of production it caused. The wording has been corrected to “a fire that shut the plant for three months” in the article body and in the closing questions list, and the same correction has been applied to the Questions page. The article’s title, summary, and every other reference already stated this correctly (“caused an outage of 3 months,” “stopped production for three months,” “suspended work for about three months”). No factual claim changes; only the loose shorthand is corrected, so the point cannot be misread as a claim that the fire burned for three months.

9 March 2026 (New Content: BBC Coverage, Leigh Day Town Hall)

Blog Posts: New

  • New post: “BBC News Lancashire Covering Hillhouse Story, and a Leigh Day Public Meeting Planned for May” (bbc-coverage-leigh-day-town-hall.md), BBC News Lancashire confirmed filming in Thornton-Cleveleys on 10 March 2026 (journalist: Yunus Mulla). Leigh Day public town hall meeting announced for late May 2026 at Thornton-Cleveleys FC. Leigh Day public contact email included. Specific dates to be updated when confirmed.

8 March 2026 (New Content: FOI Refusal, Residential Testing, 1km Zone, Planning, PFAS Plan)

Blog Posts: New

  • Biomonitoring blog post updated (biomonitoring-refused-three-day-gap.md): Previously marked [UNVERIFIED]. Now updated with confirmed sources. Council Leader’s call for blood tests: ENDS Report, 3 March 2026. Formal refusal of biomonitoring: Wyre Council FOI response, 6 March 2026 (WhatDoTheyKnow.com). Exact wording from FOI: “not appropriate,” results “uninterpretable,” would cause “additional distress.” Analysis reframed to remove editorial language while preserving the documented three-day sequence. [UNVERIFIED] tags retained only for items still unconfirmed (UKHSA guidance consultation, coordination between leader and FOI response).

  • New post: “The 1km Zone: Three Definitions, No Map, No Measurement Point”, Documents three inconsistent formulations of the advisory zone across official council documents; no published map; no measurement methodology; FOI Q5 redirected to Environment Agency. Includes a “What We Are Not Claiming” section. All three formulation citations marked as pending document-level confirmation.

  • New post: “Residential Soil Testing: 71 Properties, 7 Designated High Risk”, Documents residential testing programme: 71 properties tested, PFOA in majority, 7 designated high risk, no residential Part 2A. Covers planning moratorium: government declined Wyre’s request; council has existing LPA powers under Town and Country Planning Act 1990. Removes inference about named official’s knowledge.

  • New post: “The UK’s First PFAS Plan: Published, Challenged, and Being Watched”, Defra PFAS Plan (3 February 2026); Emma Hardy MP’s parliamentary challenge on Environment Act 2021 “polluter pays” compliance [UNVERIFIED, Hansard reference needed]; Leigh Day tracking statutory compliance. Emma Hardy section clearly marked [UNVERIFIED].

Content Updates: Existing Pages

  • health.md: Added FOI biomonitoring refusal (6 March 2026), exact wording added, WDTK link, documented divergence with Council Leader’s public call noted.
  • faq.md: Blood testing Q&A updated with formal FOI refusal; added international comparator note. Actions section updated with 71 properties / 7 high-risk stat; planning moratorium note added; biomonitoring refusal wording added. Disclaimer date updated to March 2026.
  • take-action.md: 1km zone “About the boundary” box expanded with three inconsistent formulations, no-map note, FOI Q5 redirect. Parliamentary representation updated with Defra PFAS Plan and Emma Hardy MP item [UNVERIFIED].
  • the-numbers.md: New section added under Phase 3 residential data: 71 properties tested, PFOA in majority, 7 high-risk, no residential Part 2A.
  • about.md: “Last updated” changed from February 2026 to March 2026.

6 March 2026 (Navigation Restructure, Content Updates, Part 2A)

  • Reduced from 13 top-level nav items to 6. The previous navigation bar had grown unwieldy. Reorganised into:
    • Home
    • The Facts (dropdown: What is PFAS, The Numbers, Regulatory Context, Evidence, Sources)
    • Health
    • Take Action (dropdown: Information & Resources, Resources, FAQ)
    • About
    • Contact
  • Changelog moved to footer only (was a top-level nav item).
  • Dropdown menus work on desktop (hover) and mobile (tap to expand).
  • Active state propagates to dropdown parent button when a child page is active.

Duplicate Title Headings Removed

  • The Hugo template already outputs a styled H1 from the page frontmatter title. Every content page also had a redundant H2 at the top repeating the title. Removed from: Evidence, FAQ, Health, Resources, Sources, Information & Resources, The Numbers, Regulatory Context, What is PFAS.

Part 2A Contaminated Land Classification (5 March 2026)

  • Occupation Road Allotments: Environment Agency concluded this site meets statutory criteria for formal classification as Contaminated Land under Part 2A of the Environmental Protection Act. Based on PFOA found in both soil samples and produce. If formal determination is made, Wyre Council will “identify those responsible and secure remediation.” Added to take-action.md, faq.md, evidence.md, untold.md, and homepage.
  • Sandringham Avenue Allotments: Closed precautionarily pending scientific interpretation of soil results.
  • Primary source: Wyre Council, 5 March 2026. All previous [VERIFY] tags on allotment closures replaced with this URL.

Root Uptake Science: Wash and Peel Caveat

  • Added to take-action.md and faq.md: The EA’s Contaminated Land classification was based on PFOA found within vegetables (not only surface soil). PFAS contamination in produce results from root uptake, plants absorb these chemicals as they grow. Washing and peeling removes surface soil but cannot remove PFAS already absorbed into plant tissue. The FSA’s continued “wash and peel” advice for garden produce, given this produce finding, has not been publicly explained.

FSA Advice Updated (5 March 2026)

  • New FSA recommendation added: residents within 1km should consider raised beds with fresh soil when growing produce in gardens or allotments.

Permit Compliance Failures (September 2025)

  • Added to the-numbers.md and untold.md: September 2025 compliance data recorded dioxins at 3.7× permit limit, particulates at 47× permit limit, and thermal oxidiser offline 43 days. The permit has no PFAS monitoring requirements on any of its 15+ air emission points.

W1 Shared Discharge Point

  • Added to evidence.md, the-numbers.md: W1 discharge point (20,624 ng/L PFOA) is shared by AGC Chemicals Europe and Victrex PLC (Dalmijn et al., 2025). The figure reflects combined output from both operations and cannot be attributed to a single operator without further investigation.

    [Superseded 2 August 2026. The premise of this entry was wrong. The 20,624 ng/L peak was not at W1. Dalmijn et al. (2025) record W1 as the lowest of every point sampled, at 30 and 33 ng/L total PFAS. The entry is left in place as a record of what was published at the time; the correction is at the top of this changelog.]

Historical Discharge Discrepancy

  • Added note to the-numbers.md and evidence.md: Two figures appear in the public record, 178,000 kg (Guardian/Watershed FOI, Sep 2024) and approximately 145,000 kg (EA estimate in Dalmijn et al., 2025). The discrepancy has not been explained.

Egg Testing Results

  • Added to evidence.md: Wyre Council-commissioned testing found 83 µg/kg in one duck egg [figure from ENDS Report and BBC, not yet independently verified]. Leigh Day-commissioned testing at four properties within 1km found all four exceeded the EFSA safe limit of 4.4 µg/kg (17 February 2026).

Health: Blood Testing Call

  • Added to health.md: Wyre Council leader publicly called for PFAS blood testing for residents and a “proper inquiry” into local health (ENDS Report, 3 March 2026).

Homepage

  • Added new stat: first Part 2A contaminated land classification (Occupation Road Allotments, 5 March 2026).
  • Context box updated to reference 5 March 2026 development.

6 March 2026 (Light Theme Default, Content Update)

Light Theme: New Default

  • Switched to light theme as default. Dark theme is still available via a toggle button in the navigation bar (🌙 Dark / ☀ Light). Preference is saved in your browser and remembered on return visits.
  • Rationale: this is a public health information site for a broad community audience. Light theme is more readable for general use and carries greater institutional clarity.
  • All colours verified against WCAG AA accessibility standard (4.5:1 minimum contrast ratio).
  • Dark theme fully preserved as CSS overrides, no content or layout changes.

Content Updates

  • Drop-in sessions added to Information & Resources page: 12 and 17 March 2026, Thornton-Cleveleys FC, Gamble Road, 4-7 PM.
  • Allotment closures (5 March 2026) added to current official advice section. At least two Thornton allotments reported as closed following soil test results (Blackpool Gazette, 5 March 2026, pending Wyre Council confirmation).
  • How the guidance has changed, new table added to Information & Resources page documenting the progression from “wash and peel” (July 2024) to allotment closures (March 2026), with sources.
  • 1km boundary caveat clarified in Information & Resources: no publicly available scientific justification for the 1km radius has been identified; contamination detected airborne at 20km; residents beyond 1km should not assume they are unaffected.
  • Phase 3 residential soil data (February 2026) added to The Numbers page: first confirmed residential property results (WSP code SP510); all 25 samples exceeded Dutch benchmark; peak 119 µg/kg at 0.6m depth, still rising when sampling stopped; depth-increasing pattern consistent with groundwater migration.

25 February 2026 (UI, Cancer Stats, and Content Update)

UI and Readability Fixes

  • Light-background contrast (evidence.md, health.md, the-numbers.md): Eight inline-styled divs used light backgrounds (#f5f5f5, #f0f7ff, #fff5e6, #e8f4f8) which caused white text on pale backgrounds, impossible to read on the dark theme. All converted to dark-theme-compatible semi-transparent equivalents:
    • Grey note boxes → rgba(255,255,255,0.06) with neutral border
    • Amber/warning note boxes → rgba(204,122,0,0.12) with amber border
    • Blue information boxes → rgba(0,102,204,0.10-0.12) with blue border
  • Map caption colour (the-numbers.md): Caption text using color: #666 (insufficient contrast on dark background) changed to #8b9499 for improved readability.

Cancer Statistics: Tone

  • health.md (Local Cancer Prevalence section): The 35-43% figure was previously presented in a red-bordered .stat-item.difference box, visually alarming framing. Replaced with a plain paragraph embedded within the caveat context. Added explicit statement: “Without the standardised incidence rates, it is not possible to assess whether age or other factors account for this difference.” Retained the caveats-first ordering established in the February 18 rewrite.
  • evidence.md (BBC Cancer Statistics entry): Expanded caveat to note that Thornton-Cleveleys has an older-than-average demographic and that unadjusted rates in such areas will naturally exceed national averages. Removed “above national average” as a headline finding; reframed as “reported as being above the national average” with caveats integrated. Noted that underlying standardised incidence rates were not published in the BBC article.

Content Updates (resources.md)

  • Added CHEM Trust Hillhouse-specific articles: Three articles directly covering the Hillhouse/AGC situation (Feb 5 2026; 2025 vegetables warning; 2025 PFAS hotspots feature)
  • Added Guardian Feb 2026 article covering the UK PFAS action plan
  • Added UK Government PFAS Action Plan (February 2026) to official resources
  • Added UK Environmental Audit Committee PFAS Inquiry (2025), includes note on Dr Megson’s written evidence (PFAS0073)
  • Added Italian Miteni criminal case (2025) to legal information, first successful criminal prosecution of PFAS pollution in Europe; 11 executives sentenced, 350,000 people affected
  • Added France PFAS ban in clothing and cosmetics (February 2026), first country to legislate such a ban
  • Updated CHEM Trust organisations entry to reference Hillhouse-specific coverage and EAC inquiry submission

18 February 2026 (Fact-Check Fixes)

Following a three-part fact-check review (source verification, logic & consistency, hostile reader analysis), the following fixes were applied:

Source Verification Fixes

  • Cancer statistics (health.md): Added explicit caveat that the 4.42-4.58% figures “do not appear in the BBC article and the original dataset has not been independently verified”, moved caveats BEFORE the statistics to ensure readers see context first
  • 178-tonne sub-figures (the-numbers.md): Added “[breakdown figures from EA permit documentation, specific permit version/FOI document to be confirmed]” to each breakdown figure (49,100 / 54,300 / 6,200 / 5,100 / 63,300 kg)
  • C&EN citation (evidence.md): Made more specific with link to Chemical & Engineering News and note “[article title/author to be confirmed]”
  • Guardian article date (the-numbers.md): Added specific link to 24 September 2024 Guardian article for 178-tonne figure
  • DuPont Parkersburg figure (the-numbers.md): Added “[source to be confirmed]” to the ~13,000 ng/L figure
  • $671M settlement (resources.md): Added New York Times source link for the DuPont settlement figure
  • EU Directive 2020/2184 (the-numbers.md): Added direct EUR-Lex link
  • Stockholm Convention (untold.md): Added source link for PFOS 2009 restriction claim
  • EEA-NH4 testing claims (untold.md): Added source attribution notes and corrected “soils” to “water samples” per Megson 2024 paper
  • Zone sizes (the-numbers.md): Added “[specific regulatory documents for each country’s zone sizes to be compiled]”
  • Area calculations (the-numbers.md): Added “[Area calculations derived from mapped boundaries, methodology available on request]”
  • “thousands of chemicals” (what-is-pfas.md): Changed to cite OECD figure (4,700+) with link, noting estimates vary
  • Lead/asbestos ban dates (what-is-pfas.md): Corrected lead to “Phased out 2000” (not 1999), added specific legislation links
  • No cancer studies claim (what-is-pfas.md): Added “[based on review of ECHA database and EA risk evaluation]”

Logic & Consistency Fixes

  • EU drinking water comparison (the-numbers.md): Added prominent note box explaining: measurements are from discharge water not drinking water; EU standard applies to treated drinking water at tap; comparison used because it’s the most widely understood benchmark; discharge enters watercourses where diluted before abstraction
  • Comparison table footnote (the-numbers.md): Added asterisk note clarifying “discharge water, around 206× EU drinking water standard” and explanatory footnote below table
  • Evidence.md, faq.md: Clarified EU drinking water comparison throughout, explicitly noting these are discharge measurements, not drinking water
  • Zwijndrecht zone claim (take-action.md): Added “[source for specific zone changes to be confirmed]”
  • Computer modelling claim (take-action.md): Added “[based on review of publicly available documents as of February 2026, if such modelling exists in non-public documents, we are not aware of it]”

Hostile Reader Fixes

  • Cancer statistics positioning (health.md): Moved entire caveat box BEFORE the statistics, not after, ensures readers see limitations before numbers
  • Cancer statistics (evidence.md): Reworded to lead with caveats: “Important caveats: This is third-hand data…”
  • Comparison table context (what-is-pfas.md): Added note: “This table shows regulatory timeline; it is not intended to imply toxicological equivalence between these substances”

[source to be confirmed] Tags Retained

The following tags are intentionally retained, they show transparency about what we don’t yet know:

  • AGC’s public position on PFOS usage (untold.md, faq.md, evidence.md, the-numbers.md)
  • AGC’s public position that PFOA was not manufactured (untold.md)
  • Specific EA document for EEA-NH4 testing exclusion
  • Guardian article specifics for Megson method development

18 February 2026 (Factual & Readability Pass)

Factual Fixes

  • Removed unsupported claims:

    • Removed “Hazard known 1960s (internal company research)” from what-is-pfas.md, imputes corporate knowledge without citation
    • Removed 3M Zwijndrecht 68,900,000 ng/L peak PFOA figure from the-numbers.md comparison table, extraordinary claim without verifiable source
    • Removed Megson “giant chunk of toxic material” quote from untold.md, too inflammatory for factual resource
    • Removed European site financial impact figures (€571M, €70M+) that lacked specific verifiable sources

  • Reframed unverified claims as information gaps:

    • Cancer prevalence statistics (35-43% above average) now include explicit caveat noting: data is third-hand (NHS → Ribble Rivers Trust → BBC), age standardisation unknown, no causal link established, and ward-level data is an information gap
    • “Has not been publicly explained” → reframed as “has not been determined” (neutral language, no implication of concealment)

  • Fixed editorial URL aliases:

    • Removed /what-they-dont-tell-you/ alias from untold.md (retained /the-untold-story/ for link stability)

  • Date correction:

    • Guardian article date corrected from “Feb 28, 2023” to “Feb 24, 2023” in resources.md (matches actual URL)

Readability Improvements

  • Unit definitions added throughout:

    • ng/L explained as “nanograms per litre” with EU comparison (e.g., “20,624 ng/L, around 206 times the EU drinking water standard”)
    • µg/kg explained as “micrograms per kilogram”
    • Added unit guide box to the-numbers.md

  • Technical terms defined on first use:

    • EEA-NH4: “a fluorochemical used as a replacement for PFOA in manufacturing”
    • Biomonitoring: “blood and urine testing to check for chemicals in the body”
    • Reprotoxic: “suspected of harming fertility or unborn children”
    • Non-targeted analysis: “a screening technique that looks for any chemicals present”
    • Organofluorine: “a broader category of fluorine-containing compounds, including PFAS”
    • Atmospheric deposition: “chemicals that settled from the air over decades”
    • Precursor compounds: “chemicals that can change into PFOS in the environment”
    • Ulcerative colitis: “a bowel disease”
    • Pre-eclampsia: “a serious pregnancy condition involving high blood pressure”
    • Environmental permits: “legal documents that set the rules for what a company can discharge”
    • Peer-reviewed: “studies independently checked by other scientists before publication”

  • Health page restructured:

    • Added plain-English summary box at top with key points
    • Moved caveats after facts (GOV.UK content design pattern)
    • Simplified language throughout

  • Long sentences split:

    • Multiple compound sentences over 25 words broken into shorter, clearer sentences across all pages

  • Passive voice reduced:

    • “has not been publicly explained” → “has not been determined”
    • “has been detected” → “researchers detected” / “testing found” where appropriate

  • European comparison table:

    • Added prominent context box explaining that different sites had different contamination pathways
    • Made clear that direct comparison may not be appropriate

  • Attribution language clarified:

    • “unattributed” → “source not identified” (plainer English)
    • Consistent use of “open question” for genuinely unresolved issues

18 February 2026 (Afternoon)

  • Reframed entire site as Hillhouse estate-centric rather than AGC-centric
  • Added explicit attribution for data from operator regulatory submissions (e.g., “According to AGC Chemicals Europe’s Environment Agency permit…”)
  • Distinguished between: contamination attributed to specific operators via their own disclosures, contamination detected but source unattributed, and gaps where no investigation has occurred
  • Added historical context section to About page acknowledging multiple operators (ICI 1950s-1999, AGC 1999-present) and shared estate infrastructure
  • Noted that PFOS in eggs is an open question, AGC states they never used PFOS; source unattributed
  • Added attribution methodology section to Sources page
  • Updated all references from “AGC site” to “Hillhouse industrial estate” where appropriate
  • Preserved all factual data, sources, and HTML structure

18 February 2026

  • Complete site rewrite: all content revised to factual, research-based tone
  • Removed editorial language, rhetorical framing, and advocacy messaging
  • All claims now sourced or marked for citation
  • Restructured action page as information/resources page
  • Fresh start for site content