Brought to you by Evert Bouman, Daniel Moran, Gaylord Booto, Miguel Las Heras Hernandez from NILU and  Berend Mintjes from Leiden University and A.M. Esfar E Alam from University of Oslo

Challenges in footprinting

Life-cycle assessment (LCA) is the mainstay method for environmental assessment of products. In the LCA approach, the environmental impacts of every stage in a production process, together with their upstream supply chains, are collated and a detailed environmental footprint of that product is computed. LCA concepts underpin many current green label schemes including Product Environmental Footprint (PEF), and Environmental Product Declarations (EPDs) initiatives.

Best practice dictates manufacturers perform a rigorous LCA on each product they offer. However, these studies are labour-intensive and time-consuming (read: expensive) and highly specific to individual cases (read: prone to uncertainty). Despite enormous efforts to promote standards (including an ISO series and product category-specific methodology rules) more often than not, two LCA consultants will achieve different results for the same product.

This divergence is due to the fact that LCA methods have poor quality for more remote portions of the upstream supply chain. Operations by the manufacturer in question, as well as nearby operations in the value chain or life cycle, are often well known . Contrasting, more distant portions of the supply chain are often not well known and documented only by data which is irregular and sometimes outdated (when pulled from commercial life-cycle databases) or coarse (when pulled from global supply chain models).

Development of a pan-EU or global shared data system for LCAs is a widely shared vision in the field. Such a system would enable interlinking of LCAs interwoven with global supply chain models derived from national statistical agencies. This integration would provide more up-to-date information, fill in gaps where no firm-specific supply chain information is available, and help ensure global consistency. The hybrid LCA concept enables this.

In technical terms hybrid LCA refers to the integration of a process-based life-cycle inventory with an input-output (IO) table. IO tables document supply chains at the national and international level, but at relatively coarse resolution, differentiating only 25-400 types of products for an entire economy.

Both product-level LCAs and input-output models, used to estimate product value chains have serious limitations which constrain the precision, accuracy, and depth of results. These limitations create uncertainty in results, making results less trustworthy at a time when companies and regulators are asking clearly for trustworthy footprint and value chain information. These limitations are:

• LCA and input-output models are ultimately simplified representations of complex production processes.
• Models consider average goods and are rarely linked to private company-specific or lot-specific data.
• Gathering and collating data is technically challenging.
• Data protection concerns prevent sharing. Detailed trade and production data are often closely guarded trade secrets.  
• Data is often outdated

Due to these issues, a best-in-class producer may struggle to get an LCA study to show any differentiation between their product and that of others.

The CE-RISE INFORMATION SYSTEM: A Hybrid LCA Solution

The CE-RISE Information System (CE-RISE) operationalises the hybrid LCA concept. The project aims to collect and merge firm-specific supply chain information with background information from national and international sources.

CE-RISE will capture primary information at the product, component, and material levels, and store this information in a linked environmental accounting system. This system will be integrated with recent economic accounts and other relevant statistics (for example, up-to-date electricity mixes), allowing the system to use regularised calculation methods, and provide consistent product assessment. The goal is to address the issues of data availability, recency, and comparability.  

By building on, and extending, digital product passports (DPPs), the CE-RISE information system facilitates the sharing of life cycle inventory (LCI) information across the value chain, independent of the immediate economic transactions between suppliers and buyers. This enables organisations to access primary data from a tier-n supplier, enhancing transparency and traceability. As a consequence, one can extract environmental accounts for all steps in the life cycle with the same resolution as the linked DPPs, and combine this with generic, but sufficiently detailed and up-to-date background data.

Protecting private company data is a cornerstone of the system. CE-RISE employs a blockchain based implementation of the DPP through its partner Circularise. This is expected to provide a secure framework where sensitive LCI data cannot be traced back to individual suppliers, thus maintaining confidentiality.

At NILU we are implementing the approach described above. By utilising DPPs provided by Circularise, we are constructing life cycle models. One technical challenge was the construction of the life cycle model -or value chain- represented by DPPs and automatically linking this to background datasets. The recent availability of AI language models has been instrumental in addressing this challenge. These tools enable us to bridge the gap between varied natural language used in DPPs and the standardised sector and process classifications in existing databases. This capability, together with dedicated footprints calculation algorithms, allows us to compute environmental footprints for different products, sourcing information from DPPs.

Looking ahead

We aim to scale and test this approach towards a higher quality result providing enhanced product comparability and a level of resolution superior to that of current LCA models. Automatically matching information from company-specific DPPs with background environmental-economic datasets will significantly reduce the time needed to produce footprint analyses. The approach also allows the CE-RISE system to include better and more recent information in distant parts of the supply chains, beyond what the DPP creators themselves have ready insight into. These innovations can increase accuracy and recency and decrease the labour cost of performing an LCA – a win-win for everyone.