EU Drinking Water Directive: Impact on the rubber industry

NSF senior manager, global water programs, EU and UK schemes, Samantha Duffy explains how the revised EU Drinking Water Directive (related ERJ report) will affect elastomer compounders, component manufacturers, raw material suppliers and OEMs:

ERJ: Is the revised DWD regulation set in stone: is there still scope for late revisions/derogations?

SD: The framework itself is settled. Article 11 of the Drinking Water Directive was implemented through six legal acts, three Implementing Decisions and three Delegated Regulations, adopted in January 2024 and published that April. They apply from 1 January 2027. There is no proposal on the table to reopen them, and there is no derogation mechanism that lets an individual manufacturer opt out or arrive late.

What is not fixed is the content of the positive lists. Those were always designed to be living documents. Since January this year, economic operators have been able to notify the European Chemicals Agency (ECHA) of an intention to support a substance, and from 31 December they can apply to add, amend or maintain entries. Every entry carries an expiry date: 2028, 2031, 2034 or 2037, and any entry without a review application is simply removed.

So, there is flexibility, but it sits at the substance level, and it runs through ECHA. It does not sit at the deadline. My advice is to plan for the dates as published, because waiting for a reprieve is not a strategy.

ERJ: What does this mean for seals and other related rubber/elastomer product manufacturers?

SD: Rubber sits squarely in the “organic materials” category, alongside plastics, coatings, lubricants and silicones, and organic materials carry the heaviest evidence burden in the framework.

Three things change. The first is formulation transparency. Every starting substance has to trace back to an entry on the EU positive list, not just the base polymer, but the cure system, activators, antidegradants, plasticisers, process aids and fillers. Where a substance is not listed, you need justification or additional testing. That is a supply chain data exercise before it is a laboratory exercise.

Second, testing. Migration under cold and, where relevant, warm or hot conditions; a gas chromatography-mass spectrometry (GC-MS) screen for unexpected substances, which in a cured rubber picks up reaction and degradation products as well as what you deliberately added; total organic carbon; odour, flavour, colour and turbidity; and enhancement of microbial growth.

Third is conformity assessment, EU type examination, production control with annual site inspection, an EU Declaration of Conformity and mandatory “Suitable for Drinking Water” marking. That reaches compounders, component manufacturers, raw material suppliers and OEMs alike.

ERJ: What if companies do nothing in response to the deadline?

SD: There are two very different outcomes, and which one applies depends on a single question: do you hold a valid national certificate on 31 December 2026?

If you do, such as a UBA listing or an ACS attestation, you can continue selling that product under that scheme until the end of 2032. But the cover is narrower than people assume. It is frozen: any change affecting composition or performance ends it. And a national approval remains valid only in the country that issued it, so it does not give you access to the wider EU market.

If you do not hold a national certificate, there is no transitional cover at all. From 1 January 2027, EU Minimum Hygiene Requirements (EU-MHR) conformity is the condition for placing that product on the market.

In practice, the commercial deadline arrives well before the legal one. OEMs and utilities are already writing EU-MHR conformity into specifications for post-2026 delivery. A seal supplier who cannot answer a customer’s bill-of-materials questionnaire will be designed out long before 2032.

ERJ: Which rubber/elastomer materials & products are likely to be most impacted and why?

SD: Two factors drive it: formulation complexity and whether its application is a wetted surface.

On formulation, heavily compounded elastomers have the most to prove. A typical ethylene propylene diene monomer (EPDM) or nitrile butadiene rubber (NBR) drinking-water compound carries a long ingredient list, and every one of those starting substances has to trace to a positive-list entry. Simpler chemistries have less exposure. Rubber also has a particular characteristic: because it is cured, the framework treats it as passing through an intermediate product stage, and the screen for unexpected substances captures cure by-products and degradation products, not only what you formulated in.

On wetted surfaces, classification is driven by surface area, contact time and function. Hoses and flexible connectors are treated as pipes, which is the highest risk group, so elastomer and rubber-lined hoses carry the fullest testing scope. Expansion membranes and diaphragms also sit high because of their wetted area. Static seals and O-rings inside an assembly would typically fall into the lower risk groups, with reduced requirements. But assemblies follow the pathway of their highest-risk component, so one unsupported seal can pull an entire valve or meter onto a heavier route.

ERJ: Which Member States are ahead of the legislation and which are behind?

SD: I like to separate these two questions: how ready the national administrations are, and how ready the manufacturers in each country are.

On manufacturers, those who have worked through mature national schemes are furthest along, such as Germany with UBA, France with ACS, and the Netherlands, Austria and Denmark. In our experience, the overlap with EU-MHR is substantial, so those data packages are a real head start, though the test methods and acceptance criteria are not identical and nothing transfers automatically. For example, testing with chlorinated water might be an unfamiliar process for many labs.

The tougher step is for manufacturers in Member States which don’t operate a national hygiene approval scheme. They have no certificate to hold on 31 December 2026, which means no transitional cover to 2032, and they move to EU-MHR on day one. That is counter-intuitive, and it is the group we most often find underestimating the work.

On administrations, designation of notifying and market surveillance authorities is progressing at different speeds, and industry’s own implementation guide is candid that not every Member State is likely to have a notified body. That need not be a blocker: manufacturers can use any notified body in the EU, and the certificate is valid across all 27 markets. Remember, certificates by notified bodies can only be issued from 1 January 2027.

ERJ: How big is the testing capability gap and what can realistically be done to address this?

SD: The testing capability gap is real, and it is more about duration than about the number of laboratories.

A full organic-material programme is long. Migration testing runs up to 31 days per condition. Enhancement of microbial growth runs 12 to 16 weeks depending on the method. Add formulation review, GC-MS screening, TOC and organoleptic work, and you are looking at several months of elapsed laboratory time per material before audit, technical review and certification even begin. That capacity cannot be conjured up in the fourth quarter of 2026.

Three things are compounding it. Demand is front-loaded, because no certificate can be issued before 1 January 2027. Some methods, cold-water migration under chlorinated conditions in particular, are not routine in every national scheme, so it is not simply a case of running more samples through existing benches. And more than 600 positive-list entries expire in 2028, generating a second wave of work on top of certification.

There are practical steps that can be taken now. Most compounders carry far more drinking-water grades than they need; consolidating the portfolio cuts the test count faster than any laboratory can add capacity. Screening formulations against the positive lists costs nothing but time and surfaces most failures months before a sample reaches a bench.

Proving a compound once, upstream, stops every customer re-testing it; that is what NSF Protocol 534 is for, but it lets suppliers of raw materials, pre-products and intermediates generate health-effects evidence before the regulated process starts. And finally, laboratory slots should be booked like a long-lead component. None of that buys more capacity, but it does buy back the months you would otherwise spend queuing for it.

ERJ: Any other points?

SD: Three, briefly. The first is that this is a supply chain problem, not a final product problem. A valve manufacturer’s compliance rests on evidence its seal supplier holds; the seal maker’s rests on its compounder; the compounders on their raw material producer. The framework requires formulation transparency along that whole chain, and that is a commercially uncomfortable conversation the rubber industry has largely been able to avoid until now. The companies that resolve it early will become preferred suppliers.

The second is not to read 2032 as the deadline. It is a run-off for products that already hold national approval, and it is frozen - change the compound and the cover ends.

The third is more encouraging. Once you are through it, one certificate covers 27 markets. For a seal manufacturer who has historically maintained ACS, UBA and other national approvals in parallel, the end state is genuinely simpler. The difficulty is the transition, not the destination, though EU-MHR carries no automatic recognition in the UK or North America, so global suppliers still need to plan across all three.

ERJ: Many thanks for your really helpful & insightful replies.

Image source: Freudenberg