A carbon-footprint declaration under Article 7 of the EU Battery Regulation is not a figure you look up or estimate. It is a structured result, built from a specific and prescribed body of data, and the hard part is rarely the final calculation. It is assembling — at the granularity the rules demand, to a standard a notified body will accept — data that most organisations do not yet hold in that form.
The challenges below are not hypothetical war stories. Each one follows directly from the text of Regulation (EU) 2023/1542, the JRC's calculation rules, or the draft delegated act. They are foreseeable from the rules themselves, which is exactly why they are worth planning for now.
A note on timing first, because it matters. The declaration's application date is tied to the entry into force of two pieces of secondary legislation — a delegated act setting the methodology and an implementing act setting the declaration format. As of mid-2026 the EV methodology act remains in draft; it has not been adopted or published in the Official Journal. So the formal obligation has not yet started, and the regulation gives roughly twelve months from adoption before it bites. The point is not that the deadline has passed — it is that the data-collection work below takes far longer than twelve months, so the preparation is what is urgent, not the paperwork.
The footprint is plant-specific and model-specific
Article 7(1) requires the declaration to be drawn up "for each battery model per manufacturing plant." The JRC rules reinforce it: a company-specific dataset must refer to a specific battery model produced in a specific production plant.
That single phrase has a large data-collection consequence. The same model built at two plants needs two studies, because plant-level inputs — above all the electricity mix, but also process efficiencies — differ. You cannot measure one site and reuse the result for a sister site making the identical product. The unit of work is the model-plant pair, not the model. For a manufacturer with several lines across several sites, the number of distinct studies multiplies quickly, and each needs its own primary data.
You cannot fall back on default data for the parts that matter most
For the carbon-heavy core of the battery — active materials, electrodes, cell production, and module and pack assembly — the JRC methodology requires company-specific primary data. Industry-average database values are not permitted for these processes. Secondary datasets are allowed (and in fact mandatory) only for the less significant, upstream processes.
This is the opposite of how a lot of corporate carbon accounting works, where database averages do much of the heavy lifting. Here, the processes that dominate the footprint are precisely the ones where you must measure and document real, plant-level activity data. That is where the collection effort concentrates, and it cannot be shortcut.
Much of the data sits with suppliers you do not control
The system boundary spans the whole life cycle bar the use phase: raw-material extraction and processing, manufacturing, distribution, and end-of-life. A great deal of that — cobalt, lithium, nickel, graphite, precursor and active-material production — sits upstream, with suppliers who often treat the underlying data as commercially sensitive.
This is not a theoretical tension. Independent EU-funded research (the CEPS analysis produced under the Horizon Europe BATRAW project) found, from interviews with industry, that confidentiality concerns are a key reason suppliers hesitate to share carbon-footprint data; that even when they do share figures, they sometimes withhold how those figures were derived, which complicates verification; and that data often arrives in inconsistent units, or rests on supplier assumptions that may not reflect reality.
The regulation anticipates this. The methodology provides for a neutral third party to aggregate suppliers' data into compliant datasets without exposing each supplier's raw inputs to its competitors, while still giving the notified body what it needs to verify. The point for planning is that supplier data is a negotiation and a process, not a download — and it is best started early.
The methodology is prescribed, so a generic carbon tool will not do
The calculation must follow the EU Product Environmental Footprint method, with a fixed functional unit (one kWh of total energy delivered over the battery's service life), fixed system boundaries that exclude the use phase, the Circular Footprint Formula for recycled content and end-of-life, and a specific impact method. The electricity rules are stricter than standard practice: the draft act limits what counts to the national average mix or directly contracted, physically connected supply, and removes credit for excess on-site generation.
The implication is practical. An off-the-shelf greenhouse-gas-accounting approach — even a competent one — will not produce a compliant declaration, because the boundaries, allocation rules, electricity treatment and functional unit are all dictated. Data has to be collected in a structure that maps to these requirements from the start. Collecting it first and reshaping it later tends to mean collecting it twice.
It is a living figure, not a one-off
The footprint is tied to a specific bill of materials and a specific energy mix. The rules treat a material change — a new supplier, a reformulated cell, a different material origin, a changed energy source — as producing, in effect, a new battery model that needs a fresh calculation and a fresh declaration. The draft act frames the trigger around a change in emissions beyond a set threshold.
So the data-collection task is not a project that closes. It is a pipeline that has to stay current, and detect when accumulated small changes cross the line that forces a recalculation. An organisation that treats the declaration as a one-time deliverable will find it out of date the first time a supplier changes.
Everything collected has to clear verification
The declared figure must be verified by a notified body before it can accompany the battery; it cannot be self-certified. The methodology requires a detailed, confidential supporting study containing all modelling assumptions and data-collection tables, with on-site assessment for parts of it.
That sets the standard the data must meet from the moment it is gathered: not "good enough to estimate," but "good enough to be audited." Provenance, units, assumptions and justifications need to be captured as you go. Gaps and undocumented assumptions do not disappear — they surface at verification, which is the most expensive and latest possible point to discover them.
And it has to end up as structured data, not a document
From 18 February 2027 the carbon-footprint information must be reachable through the battery passport's QR-linked record, as structured, machine-readable data alongside the rest of the Annex XIII content. Until then the declaration accompanies the battery. Either way, the destination is queryable fields in a record that can be served and submitted — not solely a report.
The planning consequence is to architect the data so it can flow into those structured fields, rather than living in a document that has to be re-keyed at the end. A workflow built around a final PDF creates rework; one built around structured data does not.
The through-line
None of these challenges is exotic, and none is invented. Each follows from the rules: model-and-plant granularity, mandatory primary data for the core, upstream data you have to negotiate for, a prescribed methodology, recalculation when inputs change, audit-grade evidence, and a structured destination. The organisations that find the declaration hardest will be the ones that treat it as a calculation to run once the act lands. The ones that find it manageable will have treated it as a data-collection discipline to stand up in advance — gathered at the right granularity, against the right method, kept current, and built to be verified and served.
That is the real work, and it is available to start now, regardless of the date the act is finally adopted.
Work the numbers first. We have published a free carbon footprint worksheet: enter the four JRC CFB lifecycle stages, get the declared intensity in kg CO₂e/kWh, and see exactly which passport field each figure belongs in.
Wat er is veranderd in de EU-regelgeving voor batterijen en productpaspoorten, wat dat betekent voor operators en de data die eraan komen.
Zet de volgende stap
Klaar om conform te zijn op 18 februari 2027?
EU Digital Passport Processor maakt, host en dient EU-batterijpaspoorten in voor fabrikanten en importeurs. Demo-accounts zijn op aanvraag beschikbaar.