The role of the PCR in EPD Automation
Presented at ACLCA 2025 in Atlanta, GA
Suddenly, EPDs are everywhere.
Procurement. State agencies across the country (among them Minnesota and California) are piloting programs to require ISO 14025 environmental product declarations (EPDs) in procurement processes. This follows on an ambitious but short-lived Federal incentive program.
Informed Consumption. The concept of presenting EPD data as a sort of environmental “nutrition facts” label on consumer products is perennial among startups (see this one, for example).
Automation. Other startups are focusing on AI-driven automation to accelerate the production of EPDs
However, there are a few issues:
LCA at Scale is required to generate EPDs for the huge variety of different products available for purchase.
That means working quickly. The ordinary 6 to 12-month development cycle for LCA studies leading to an EPD is simply not going to cut it.
But most of all, EPDs require third party verification. Rapid, widespread deployment of auto-generated EPDs, the dream of LCA practitioners for years and decades, could turn into a nightmare if current practices for verification are not improved.
“Verification at scale” is not a thing.
People in the business of making automated EPD generator tools will tell you there’s an easy answer to this problem: it’s called “pre-verification” and it means “looking over the tool” and giving it an official, third-party expert stamp of approval before any EPDs are generated.
The problems with this?
There are no standard practices for pre-verification of LCA tools, something that was apparently not contemplated in the ISO 14025 standard.
The quality specifications provided in Product Category Rules (PCRs) are widely varying, as are the general program instructions published by EPD program operators.
For that matter, data quality practices for LCA itself are widely varying. The near-exclusive reliance on proxy background datasets means that every LCA result is a carefully-circumscribed fiction, and that may not be compatible with what the public expects in a Nutrition Label.
Nobody has yet decided what it means to “verify” EPD results generated en masse. Nobody has even decided whose job it is to decide what this means. Is it the Program Operator’s job? is it the job of the PCR Technical Committee? Is it the Verifier themselves who gets to set the standard?
One thing is for certain: “Pre-verification” Is not actually verifying anything, at least from the common-sense definition of making sure the numbers are correct. But third party verification of thousands or millions of product carbon footprints is not tenable using current practices. To remedy this situation is going to require a thoroughgoing change in the entire EPD generation pipeline, beginning with the PCR Technical Committee. But what would that look like?
Stay tuned for more tomorrow, when I talk about all the different stakeholders involved in the EPD generation process.
