Environmental Product Declaration (EPD)
A verified document quantifying environmental impacts of a material or product across its lifecycle, expressed in kg CO2e and standardized indicators.
What is an Environmental Product Declaration?
An Environmental Product Declaration (EPD) is a verified, third-party document that quantifies the environmental impact of a material or product across its lifecycle using lifecycle assessment (LCA) methodology. Unlike marketing claims, an EPD is built from measured data, follows EN 15804 or equivalent standards, and is independently verified by an accredited program operator. The primary figure reported is Global Warming Potential (GWP) in kg CO2e, though EPDs also cover acidification, eutrophication, and resource depletion. An EPD is the direct data source used to compute embodied carbon and support whole-life carbon assessments. The EPD system exists because environmental claims are often vague and unverifiable. An EPD replaces narrative with traceable, auditable numbers that can be independently checked.
What makes two EPDs comparable?
Two EPDs are only comparable if they declare the same unit, cover the same lifecycle modules, and follow the same product category rules (PCR). An EPD reports impacts across lifecycle phases: modules A1–A3 (production: extraction, manufacturing, transport to site), B (use), C1–C4 (end-of-life), and D (recycling credit). Most building material EPDs stop at A3 (cradle-to-gate). A per-kilogram figure for concrete, A1-A3, cannot be compared to a per-m3 figure, and neither can be compared to one covering full lifecycle (A1-D). A designer who reads two numbers without checking units and modules has not compared materials; they have compared different boundaries.
| Declaration | Unit | Modules | Comparable to What? |
|---|---|---|---|
| Concrete 1.2 kg CO2e/kg | Per kilogram | A1-A3 | Only other per-kg, A1-A3 concrete |
| Concrete 0.35 kg CO2e/m3 C30/37 | Per cubic meter | A1-A3 | Only other per-m3, same strength class |
| Timber 0.2 kg CO2e/kg A1-D | Per kilogram | Full lifecycle + recycling | Not with A1-A3 figures; different boundary |
Module D (recycling) can shift results by 30–50%. The PCR also matters: concrete EPDs specify cement type, reinforcement, and transport distance. Two EPDs following different PCRs are incomparable even if unit and modules match. This incomparability is not a weakness of the system; it is intentional. It forces designers to ask hard questions before reaching for the lowest number.
Why does a specific EPD sometimes look worse than an industry average?
A manufacturer-specific EPD reports measured data from one factory. An industry average aggregates multiple producers. A generic figure is a conservative database placeholder. When a specific EPD shows higher emissions than an average, it often means the manufacturer measured honestly while the average hides inefficiency through aggregation. A Slovak ready-mix plant with a fuel-efficient kiln may declare 0.28 kg CO2e/kg. A generic European figure might be 0.35 kg CO2e/kg, averaged across older, less efficient plants. The specific EPD is the better choice and reveals competitive advantage. Dismissing a higher-looking specific number in favor of a lower generic one is a common mistake and often results in unintentional greenwashing.
| Type | Source | Use When |
|---|---|---|
| Manufacturer-specific | Measured from one factory | Specifying a named material from a known producer; highest relevance |
| Industry average | Aggregated from multiple factories | Early design phase when supplier is not yet fixed |
| Generic | Database estimates | Last resort, when measured data unavailable; always disclose as generic |
What separates a verified EPD from a marketing claim?
Third-party verification is the crucial difference. An accredited verifier audits the LCA model, data sources, and calculations. The program operator (such as the International EPD System, IBU Germany, or EPD Italia) registers the EPD, assigns a unique number, sets a validity period (typically 5 years), and maintains a public registry. A legitimate EPD carries a registration number; without one, a low number is merely a company's assertion, not a verified fact. When comparing materials, verify that the program operator is listed on an accredited scheme. An unverified claim, however attractive, should not influence specification.
In Slovak projects, coverage is uneven: timber has good data, but many domestic ceramics manufacturers and smaller ready-mix plants do not publish EPDs. Acknowledge this in specifications: require EPDs from accredited operators, accept industry averages if manufacturer-specific data does not exist, and allow generic figures only if the contractor explains why measured data is unavailable. This approach is honest and reflects market reality.
Why do EPDs expire and what is the right age to use?
EPDs typically expire 5 years after publication. Production efficiency, grid decarbonization, and supply chains change. An EPD from 2019 may no longer reflect current performance. In specifications, state the maximum age you will accept: typically 2–3 years. An expired EPD should not justify material choices in a contemporary design. If a manufacturer has upgraded equipment or sourcing since an EPD's publication, the current product may perform significantly better.
How should you collect and interpret EPDs for specification?
Collect EPDs at the specification stage, before tender. This is where environmental criteria drive design decisions and materials can still be changed. After tender closes, material is already decided and your leverage is gone. In the specification, require the declared unit, minimum lifecycle modules (usually A1-A3 for materials), the program operator name, and maximum publication age. Ask contractors to declare whether figures are manufacturer-specific, industry average, or generic, and to justify why manufacturer-specific data was unavailable if applicable.
The most useful skill when reading an EPD is skepticism toward headline numbers. A figure of 0.15 kg CO2e per kg is meaningless without context: which modules, which declared unit, from which factory, verified by whom? A low number without this context is a red flag. Always verify the declared unit and modules match across materials you are comparing. A specific EPD reporting higher emissions than a generic average may be the more honest choice; measured data often reveals real costs that estimates hide. But you cannot make informed trade-offs if you compare incomparable numbers. Honest specification and honest reading of EPD data are the foundation of genuine embodied carbon reduction.
Frequently asked questions
- What exactly does an EPD tell you?
- An EPD quantifies the environmental impacts of a material or product across defined lifecycle stages (from raw material extraction through manufacturing, transport, use, and end-of-life) using methodology called lifecycle assessment (LCA). The primary figure most architects read is Global Warming Potential (GWP), expressed as kg CO2e, but an EPD also reports other impacts like water use, acidification, and resource depletion.
- Why can't I just compare two EPD numbers directly?
- Two EPDs are only comparable if they declare the same functional or declared unit (per kg, per m2, per m3), cover the same lifecycle modules, and follow the same product category rules. A per-kg figure for steel tells you nothing against a per-m2 figure for insulation. A cradle-to-gate declaration (A1-A3) that omits use and end-of-life phases cannot be compared to one that includes them.
- What are lifecycle modules A, B, C, D?
- Modules A1-A3 cover production (raw material, manufacturing, transport to plant). Module B covers the use phase (maintenance, repair, replacement over typically 50 years). Modules C1-C4 cover end-of-life (deconstruction, transport, processing, disposal). Module D captures the value of recycling beyond the product boundary. Many EPDs stop at A3 (cradle-to-gate), leaving B and C unknown.
- What's the difference between a manufacturer-specific EPD, an industry average, and a generic figure?
- A manufacturer-specific EPD reports measured data from an actual factory: it reflects that producer's efficiency and can be higher or lower than the average. An industry average (sometimes called a Type III declaration average) combines multiple producers' data. A generic figure is a placeholder from a database, often pessimistic, used when no specific data exists. A specific EPD that looks worse than an average may be honest; measured data often reveals inefficiencies marketing hides.
- How do I know if an EPD is trustworthy?
- Look for third-party verification by an accredited body and a program operator (such as EPDItaly, IBU in Germany, or the International EPD System). These operators enforce methodology, verify data, and assign validity dates (typically 5 years). An EPD without independent verification is marketing material, not a technical declaration. Check the validity date; expired EPDs should not be used in specifications.
- When should I collect EPDs in the project workflow?
- Collect EPDs at the specification stage, not after tender. At specification, you can influence material choice and drive low-carbon options. By the time bidding closes, the material is already decided, and you have no leverage. Asking for EPDs in a specification also signals to suppliers that environmental performance matters, encouraging them to source better data and materials.