EPD and PCF models for customer controlled supply chains
When customers dictate suppliers, materials, and logistics, manufacturers carry the reporting burden without full control of the inputs. The trick is not to promise x‑ray vision. It is to design an EPD and product carbon footprint model that sets defendable boundaries, absorbs customer data cleanly, and fills gaps with credible proxies while protecting IP.
Start with boundaries that match reality
Pick a reporting scope your auditors and buyers can both live with. For an EPD, anchor on EN 15804 modules A1 to A3 as the core, then state how customer‑mandated suppliers affect A1 data access. For PCFs, state cradle‑to‑gate as default, then explain which parts of upstream data are primary and which are modeled.
Tell readers exactly what is inside the gate, what is proxied, and why. That clarity lowers pushback later.
Who owns which emissions
Operations own Scope 1 and 2. Parts of Scope 3 are shared with the customer because they select suppliers, routes, and sometimes packaging. Align early on Category 1 purchased goods, Category 3 fuel and energy upstream, and Category 4 transport. Document the allocation keys you will use when primary data is missing, such as mass, economic value, or process time.
If the customer requires an allocation method, log that choice as metadata so every recalculation reproduces it.
A pragmatic data‑model blueprint
Your schema should mirror how the product is really made, not how the org chart looks. At minimum, model these objects and relationships:
- Product version and PCR reference (with revision and verification metadata)
- BOM positions linked to material datasets and supplier‑facility IDs
- Facility energy meters by vector and time slice, not just annual totals
- Transport legs with mode, distance, payload assumption, and backhaul flag
- Data quality scores per record plus a provenance trail and consent
- Legal constraints and confidentiality scopes at the field level
We prefer schemas that treat every upstream record as time‑bound and replaceable. It keeps re‑baselining fast when suppliers change.
Want to optimize your EPD data governance?
Follow us on LinkedIn for insights that help you streamline compliance, improve reporting accuracy, and unlock new project opportunities.
When to trust customer data vs fill the gaps
Accept facility‑specific EPDs or third‑party verified PCFs as primary. Next best is supplier‑specific data reviewed under a recognized method. If neither exists, move to reputable generic datasets for the material and geography, then note the substitution. Be explicit about cutoffs and coverage so reviewers know what was estimated.
If the buyer sends spreadsheets, map them into your canonical fields the same day. The faster you normalize, the sooner you see what is still missing.
Supplier outreach at scale without IP drama
Ask only for what changes the math. Typical minimums are material recipe share, energy by vector, scrap, and on‑site fuels. Also request EPD or dataset IDs, facility location, and the reference year. Keep commercial terms out of the request to reduce redlines.
Use a portal that lets suppliers upload confidentially and locks raw files behind role‑based access. A good model tracks consent and masks fields in exports, so nothing leaks into customer hands by accident.
Fast wins most teams miss
Grid electricity is often the largest controllable lever in A3. Use subregion emission factors instead of a national average when you can. The U.S. average grid intensity was about 0.39 kg CO2e per kWh for 2022 output according to eGRID, and subregion variation can materially swing results (EPA eGRID, 2024) (EPA eGRID, 2024).
Transport mode selection matters. Air freight can exceed 500 g CO2e per tonne‑km while deep‑sea container shipping can be around 10 to 40 g CO2e per tonne‑km depending on assumptions, a gap that drowns small material tweaks (UK Government GHG Conversion Factors, 2024) (UK Government GHG Conversion Factors, 2024).
Calibrate expectations with numbers
Supply chain emissions can dwarf on‑site emissions. Across reporting companies, value chain emissions have been measured at over eleven times operational emissions on average, which explains why perfect primary data is rare in early cycles (CDP, 2024) (CDP, 2024).
Use that ratio to set stakeholder expectations on the first pass vs year two refinements. No one wins by promising full cradle‑to‑gate precision in week one.
PCR fit and declarations that age well
Pick the PCR most used by peer products so comparability lands with specifiers. Note any module exclusions and the share of modeled vs primary data in the project report. If the relevant PCR is due for revision within your planning horizon, schedule data model tweaks now so renewal is a tidy update, not a rebuild.
If valid EPDs must be refreshed sooner because of PCR changes, lock your schema to keep old and new factors side by side for trend views. Dont let versioning live only in someone’s inbox.
Governance beats heroics
Great EPDs and PCFs in customer‑controlled supply chains are built on rules, not heroics. Put boundaries, allocation keys, and data provenance in the model. Automate supplier requests and protect what must remain confidential. Then iterate. The result is credible transparency buyers can act on, with less drama and more speed.
Frequently Asked Questions
Which emission categories should manufacturers prioritize first when customers mandate suppliers?
Focus on Scope 2 electricity at your facilities and Scope 3 Category 4 transport. Both often shift results quickly with better regional grid factors and optimized modes, and they can be improved without changing customer‑mandated materials (EPA eGRID, 2024) (EPA eGRID, 2024).
What is an acceptable fallback when primary supplier data is not available for an EPD or PCF?
Use reputable generic LCI datasets matched to material and geography, document the substitution in the project report, and tag those records for future replacement. Be explicit about any cutoffs and modeled shares. If the buyer provides verified facility data later, swap it in and keep both versions for audit.
How should allocation be handled when multiple products share a process and only rough data exists?
Choose a defensible rule such as mass, economic value, or machine time, apply it consistently, and record it in metadata so recalculations reproduce the same outcome. Where the customer dictates the rule, store that decision at the process level and lock it for traceability.
About the Author
Walker Ryan is a climate-tech entrepreneur focused on driving industrial decarbonization through better data. As the founder and CEO of Parq, he helps manufacturers generate high-quality, third-party–verified carbon disclosures at scale—accelerating a traditionally slow and expensive process. Before starting Parq, Walker led over $200 million in sustainability-focused investments as VP of Strategy & Growth at ReStream Solutions, following earlier experience in investment banking at Deutsche Bank. He brings a rare mix of capital markets expertise and hands-on sustainability knowledge to tackling the infrastructure of industrial emissions.
