Do I need separate LCAs for each plant?

Not always. You must **model** each plant separately, but you can combine the results into one declaration if variation stays within PCR limits (often ±10 %).

Can I use supplier-specific electricity data for my EPD?

Yes, when the PCR allows and you have audited evidence. Otherwise default national grid factors are safer.

How often should I update a multi-site EPD?

Whenever production volumes or energy mixes shift enough to push variation beyond thresholds, or at least every five years to stay within EPD validity.

What if one plant closes during the EPD’s validity period?

Publish an addendum or revise the EPD to reflect the new production mix so users do not rely on outdated averages.

Does transport mode choice matter?

Absolutely. Switching from air to sea freight can cut cradle-to-gate emissions by over 80 % for long-haul legs (NMD, 2025).

Is a worst-case EPD penalized in LEED scoring?

No. LEED looks at declared numbers, not the methodology behind choosing worst-case. However, lower values obviously score better in the whole-building LCA credit.

Multi-Site Manufacturing: One Product, Clear Carbon Math

Why geography skews your LCA numbers

A kilowatt-hour in Poland carries roughly 600 g CO₂e, while the same unit in Sweden hovers near 30 g CO₂e and nuclear-dominated France lands around 70 g CO₂e (IEA Grid Emission Factors, 2024). In the United States, regional intensity ranges from 242 to 1 549 kg CO₂e per MWh (EPA eGRID, 2025). Layer in ocean freight or cross-country trucking and two “identical” products suddenly look nothing alike on paper.

Start with a split-inventory

Build one bill of materials per site. Capture on-site fuel, electricity, water, and waste in separate columns. Keep upstream inputs identical where they truly match, but resist the urge to average early. A tidy split inventory becomes your pivot table for allocation later.

Allocate production volumes the ISO way

ISO 14044 calls for allocation that reflects actual output, not wishful thinking. Weight each site’s impacts by its share of annual production, then sum for a volume-weighted global result. If Plant A produces 70 % of annual units, its carbon footprint should steer 70 % of the blended EPD result.

Harmonize transport legs and energy grids

Choose consistent cut-off rules for inbound transport. If rail to Port X is included for one facility, include the equivalent leg for every facility. For electricity, use country-level grid factors unless your PCR permits supplier-specific data. Consistency beats granularity when auditors review your workbook.

Model scenario blends inside the PCR limits

Most construction PCRs now allow three disclosure styles: site-specific, representative average, or conservative worst-case. EPD International’s GPI 5.0 flags a 10 % variation threshold—above that, an average may mislead and separate site EPDs are recommended (GPI 5.0, 2024).

Document assumptions like a food label

List each site, its grid factor, and its transport distances in the EPD annex. Think of it as an ingredients panel for carbon. Clear traceability builds buyer trust and speeds third-party verification.

When to publish one EPD or several

One blended EPD works when variation stays below 10 % and buyers care more about portfolio breadth than pinpoint accuracy.
Separate EPDs win when a low-carbon flagship site offers competitive edge or when regulatory credits reward site specificity.
A conservative worst-case EPD can bridge gaps while a lagging facility upgrades.

Turning complexity into spec power

Managing multi-site data feels messy, yet the payoff is real. Transparent allocation lets you showcase best-in-class plants without hiding the rest, unlock low-carbon product lines faster, and meet specifiers’ appetite for hard facts over green gloss. Complexity handled well becomes one more reason your product stays on the bid list.