How Long Does an Environmental Product Declaration Take, and What Data Do You Need First?

It is usually the first practical question once an organisation has decided an EPD is worth pursuing: how long will this actually take, and what do we need to have ready before we start? The honest answer is that timing depends heavily on data readiness, not on the paperwork itself, and knowing what “data ready” actually means is the best way to avoid a project that drags on for twice as long as expected.

The main steps, and where time is actually spent

Developing an EPD generally follows five stages: selecting the applicable Product Category Rules, conducting a life cycle assessment based on that PCR, compiling the results into the standardised EPD reporting format, independent third-party verification, and finally registration and publication. In practice, the first two stages, selecting the right PCR and conducting the LCA, consume the large majority of the timeline. Compiling the reporting format and going through verification tend to move relatively quickly once a solid LCA is in hand, and once the EPD documentation and service agreement are finalised, publication itself is required to happen within 90 days of the EPD’s version date.

The variable that actually determines your timeline is data. An organisation with clean, well organised primary production data can often move through the LCA stage in a matter of weeks. An organisation starting from scratch, with data scattered across suppliers, spreadsheets, and undocumented assumptions, should expect months, not weeks, and that is before accounting for the verifier’s review cycle.

The data you actually need, and the standards it has to meet

Primary data for your own operations. For the processes you directly control, whether that is manufacturing, processing, or service delivery, primary data is required, not optional. This means your own site-specific figures on material and energy inputs, outputs, and waste, generally averaged over at least a full year of operation (which does not need to be a calendar year), and no more than five years old at the time the data is used.

Supplier data, where it is available. For upstream processes outside your direct control, primary data should still be used where available, for example site-specific figures for materials from contracted suppliers. Where genuine primary data cannot be obtained, representative secondary data from databases or literature can be used instead, provided it meets defined quality requirements: a reference year no older than ten years, geographical and technological representativeness of your actual supply chain, and use of the most current version of the relevant database.

A defined declared or functional unit. Before any data collection really gets underway, you need clarity on the reference unit your results will be expressed against, whether that is a quantity of product (a declared unit, such as one kilogram of material) or a quantified performance (a functional unit, such as one square metre of surface covered for ten years). This is generally fixed by the applicable PCR, but understanding it early shapes exactly what data you need to collect and how.

A realistic view of your data gaps. Where primary or representative secondary data genuinely cannot be sourced, proxy data can be used as a last resort, but it is capped: proxy data cannot contribute more than 10 percent of the results for any single impact indicator. If your operation has significant data gaps beyond that threshold, closing them before starting the formal LCA is far more efficient than discovering the gap partway through verification.

Why data quality drives the transparency in your EPD

It is worth knowing that the eventual EPD will need to disclose, not just report, information about the data quality behind it. For any process contributing more than 10 percent to the climate change results of the core production stages, the EPD needs to state the type and source of data used, its reference year, and whether it is primary or secondary. The overall share of primary data behind your core production results also needs to be disclosed, alongside a standard caveat explaining that this is a simplified data quality indicator. None of this is unusual or punitive, but it does mean data quality is not something to sort out after the fact. It is part of what gets published.

A realistic timeline

For an organisation with reasonably organised production data and a clear applicable PCR already in place, three to six months from kick-off to a verified, published EPD is a realistic range. For an organisation starting with limited data infrastructure, developing a first EPD can reasonably take six months to a year, factoring in the time to establish reliable primary data collection, source appropriate secondary data for the rest of the value chain, and move through independent verification. Once the first EPD exists, updates and subsequent products in the same category typically move considerably faster, since much of the data infrastructure and methodology is already in place.