EF 3.1 has expired: which LCA database for your PEF in 2026?

1The EF 3.1 database expired on 31 December 2025 and has no immediate successor

The EF 3.1 database is the common foundation of all PEFCRs published since 2018: 4,400 life cycle inventory datasets, including more than 800 sectoral datasets, developed under a European Commission mandate and hosted on the European Platform on Life Cycle Assessment. Its validity was contractually limited: the commercial providers that supplied the datasets (ecoinvent, Sphera, Blonk, Quantis and others) retain their intellectual property and signed five-year availability agreements.

These contracts ended on 31 December 2025. Since that date, EF 3.1 has no longer been available for new studies, and no official EF database has replaced it. This is neither a recall nor a methodological flaw: as the Commission reiterates in its June 2026 communication, the scientific robustness of EF 3.1 is not in question. It is purely the effect of a contractual expiry.

For any manufacturer undertaking a PEF, a DPP or environmental labelling, this creates an operational grey area: the method remains applicable, but the database used for the calculation is no longer the same as the one used by earlier studies and published sectoral benchmarks.

2The Commission's June 2026 guidance permits two commercial databases during the interim period

Faced with this gap, the European Commission published guidance in June 2026 on the transitional use of commercial databases under the PEF method. The principle: until EF 4 is published, expected in the first quarter of 2028, a PEFCR can append a list of commercial datasets that its users must use to preserve comparability between studies in the same sector. One dataset per use, to prevent two companies calculating the same impact with different databases and obtaining structurally divergent results.

The Commission's selection criteria for assessing candidate commercial databases are explicit: coverage of the datasets required by the PEFCR, a licensing model compatible with widespread use, interoperability between LCA tools (SimaPro, GaBi/Sphera, openLCA, PEF platforms), the quality of sourced data and transparency of construction methods. Two databases pass this filter for current large-scale uses: ecoinvent for core datasets (energy, transport, basic materials) and WallDB (World Apparel and Footwear Life Cycle Assessment Database, developed by Quantis) for textile-footwear sectoral datasets. Other sectors will need to carry out an equivalent exercise as their PEFCRs are revised.

3The textile PEFCR provides the first operational mapping: 641 sectoral datasets reassigned

The apparel & footwear v3.1 PEFCR, formally adopted by the European Commission in June 2025 and hosted by 2B Policy, is the first to publish a sectoral dataset mapping annex. The starting point: 641 sectoral datasets needed to cover apparel and footwear life cycles, extracted from the EF 3.1 list and reworked by the Technical Secretariat (removal of datasets with no use, addition of missing datasets identified by manufacturers).

The matching method is strict: comparison of names + metadata dataset by dataset, prioritising ecoinvent where coverage is equivalent, and using WallDB where ecoinvent has no suitable dataset. Each match is categorised into four levels:

• Perfect match: identical technology and region.
• Good match: the same technology, a nearby region.
• Partial match: identical basic technology, a different specific process.
• Proxy: a generic substitute for a missing specific dataset.

The result as of 22 July 2026: 317 datasets matched to ecoinvent, 322 to WallDB, and 2 still unmatched. A review cycle is underway: each data provider validates its own proposals, then the Technical Secretariat reviews them and suggests adjustments before publication of the final annex, announced for September 2026 during the 22 July webinar: check that it has been published before finalising your calculations.

4What this changes for your PEF, DPP and environmental labelling calculations in 2026

Three types of operational consequences are documented by manufacturers involved in the Technical Secretariats. Reduced comparability between brands in the same sector until the sectoral PEFCR mapping fixes the choice of databases. Reduced internal consistency in studies by the same manufacturer that would need to mix several databases to cover a full life cycle. Granularity gaps for datasets without a direct equivalent (for example, chemical recycling) or available at a different aggregation level, affecting application of the Circular Footprint Formula (CFF).

Three use cases raise different issues. For textile environmental labelling (the French Écobalyse scheme, with equivalent schemes in preparation for other sectors), the calculation follows updates to the public tool, which draws on its own sources (ecoinvent, Base Impacts, processes built by the Écobalyse team), rather than the PEFCR mapping. Scores calculated before and after a database change will not be strictly comparable: manufacturers must document this in their communication, or risk accusations of greenwashing over unexplained score changes.

For a DPP, the situation depends on the category. The Batteries Regulation (EU) 2023/1542 defines its own CFB-EV method through Article 7 (delegated act still pending): the EF 3.1 issue is indirect here. The other categories covered by the ESPR will await their sectoral delegated act, which will necessarily align with the EF database available on its adoption date. A manufacturer preparing its DPP in 2026 should prioritise quality primary data collected from its suppliers rather than investing in detailed mastery of a secondary database that will change by 2028.

For an internal LCA calculation for eco-design purposes, the constraint is less severe: what matters is choosing a consistent, documented and maintained database, then rerunning the same calculation with the same database to compare before/after. Only comparability within the portfolio matters here. LCA consultants help each manufacturer make this choice according to its sector and history of studies.

The fundamental change in EF 4, expected in the first quarter of 2028 for its core datasets, is a redesign of governance. The envisaged model: the European Commission becomes the owner of the core datasets' intellectual property (energy, transport, basic materials), allowing them to be maintained continuously as industry evolves (for example, the decarbonisation of the Chinese electricity mix observed between 2020 and 2026), rather than freezing a version for five years before it expires abruptly.

Sectoral datasets, for their part, would be contributed by sector federations, academic actors and commercial providers within an open governance framework, with centralised technical review. This model remains theoretical and unconfirmed as of 22 July 2026: it is the direction announced by the Commission in its workshops, but the formal act has not been published. It needs monitoring, bearing in mind that EU timelines often slip by 12 to 18 months compared with initial announcements: a cautious practitioner plans to use commercial databases during the interim period until mid-2028, rather than Q1 2028.

One, map your existing PEF or LCA studies: which were calculated under EF 3.1, which additional databases were used, and which results are used externally (labelling, customer communication, CSR report, CBAM file or DPP).

Two, decide between waiting for EF 4 (2028) and investing in a commercial database now. The right answer depends on the sector (textiles have already switched), regulatory urgency (DPP, labelling) and budget: ecoinvent and WallDB licences cost a few thousand euros per year for a single user; shared access through a consultant or a PEF-compliant tool is often more economical.

Three, make the collection of primary data from your suppliers systematic (site energy mix, transport, material allocation): these data determine the calculation's reliability and remain robust through all secondary database changes. A well-maintained Bilan Carbone® or a formalised supplier data collection system provides this foundation.

5Key takeaways

The EF 3.1 database expired on 31 December 2025 because of contractual arrangements. In June 2026, the European Commission published guidance permitting the interim use of commercial databases, with one dataset per use to preserve comparability. The textile PEFCR has adopted ecoinvent + WallDB; other sectors will follow. EF 4 is expected in the first quarter of 2028 with new governance: Commission ownership of core IP, decentralised contributions for sectoral data. In practice, map your existing studies, decide on investment in a commercial database according to your regulatory timeline, and make primary data collection systematic.

Further resources

Frequently asked questions

1Can EF 3.1 still be used in 2026 for a PEF calculation?

Not for a new study or a study for regulatory or external communication purposes: the database is no longer contractually made available. A study already finalised under EF 3.1 remains valid in the context in which it was delivered, but is not directly comparable with studies calculated using the new database.

2Must all earlier PEF studies be redone with the new database?

Not systematically. A study used internally for eco-design does not need to be rerun if it continues to guide sound choices. A study used in external communication (labelling, customer sales arguments) must be rerun if the sectoral PEFCR has published its new mapping and the version published by the competitor is calculated using the new database.

3How much does an ecoinvent or WallDB licence cost?

Ecoinvent: around 3,000 to 15,000 euros per year depending on the number of users and type of organisation, with preferential academic rates. WallDB: access negotiated through Quantis, with a general framework across multiple tools being structured according to the Commission. Shared use through an LCA consultancy spreads this cost across individual studies.

4Is the apparel & footwear PEFCR mandatory?

No, it is voluntary: a manufacturer can carry out a PEF without a PEFCR (using a more complex and costly method) or choose other frameworks (ISO 14040/14044). The PEFCR effectively becomes mandatory when a regulatory scheme relies on it. This is not the case for French textile environmental labelling: Écobalyse uses the PEF foundation, with its own adjustments, 2 additions outside LCA (microfibres, exports outside Europe) and a durability coefficient.

Ask Sébastien your questionA personal reply within 24 working hours

Send my questionor: [email protected]