MEDIA GALLERY – THE PLATFORM’S STORY, TOLD THROUGH ITS STUDIES

Every study hosted on the EPT platform belongs to the organiser who opened it.

This page is where the platform tells its own side of the story: the dossiers, the conference presentations, the videos. What a collaborative interlaboratory study can do – shown, not claimed.

INTERLABORATORY IN LCA – INTERLABORATORY STATISTICS MEETS A COMPLEX SYSTEM

The scientific question.

Life Cycle Assessment is standardised by ISO 14040/14044, yet studies in the literature had shown that different practitioners analysing the same product reach markedly different results. Those comparisons, however, lacked a statistical protocol – and a comparison without a statistical protocol permits no inference: it cannot tell significant differences from chance, so it cannot be interpreted. The variability of LCA had been observed many times; it had never been measured. The real question behind this study was therefore larger than LCA itself: can the statistical framework of interlaboratory studies, which the literature had already applied to complex systems such as energy systems, master a complex calculation system like LCA?

The protocol.

The study answered by design. The ISO/IEC 17043 / ISO 13528 interlaboratory framework was applied to LCA practice for the first time, in the collaborative form that is the EPT platform’s signature – and with a deliberate addition. Because the scientific question included the qualification of practitioners, the protocol was designed with an evaluative dimension as well: proof that on the EPT platform the protocol is redesigned, every time, around the question it must answer. Starting from more than 50 potential variability factors identified in the literature and in the standards, some factors were constrained and others left free, so that their effect could be isolated and measured. Three sequential case studies in the food and packaging sector were designed: packaging production (software and databases), a dehydration process (energy-mix modelling) and pasta production (allocation choices). The statistical expertise behind the platform also served as technical reference in the development of UNI 11698:2017, the Italian standard on the estimation, declaration and use of the uncertainty of LCA results.

The participants.

Over 80 professionals from 17 countries registered; 60 took part in the tests. Participation was free of charge and protected by the platform’s anonymity architecture – every result identified by a PIN alone. Practitioners who wished to obtain a personal qualification could voluntarily disclose their identity: an opt-in choice, managed entirely within the project page, that never affected the anonymity of the other participants.The outcome. The study produced consensus reference values for the three case studies and, for the first time in the field, the real measurement uncertainty of LCA results – quantified and separated by source: the component coming from the databases (estimated by Monte Carlo methods) distinguished from the component coming from the practitioners’ choices (estimated through the interlaboratory approach). Where the protocol constrained the practitioners’ degrees of freedom, uncertainty dropped to a few percentage points; where choices were left free, it grew several-fold. And for the first time, LCA practitioners were qualified on objective statistical evidence: each performance evaluated through z-scores under ISO 13528 – data, not opinions – opening the way to a register of qualified LCA practitioners.

The partnership.

The study was developed by the Effective Proficiency Team – the industrial-statistics expertise behind the EPT platform – together with MAPPING LCA, a group of sustainability experts including a consulting company and two university laboratories, with the Agrifood LCA Laboratory of the University of Milan. Results have been presented at the LCA FOOD 2024 international conference, at the MAPPING LCA annual conference (Milan, 2025), and in advanced training for the Italian LCA Network.

VALIDATING AN INTERNATIONAL STANDARD – OFFICE SEATING, DIMENSIONAL TESTING

The scientific question.

An ISO technical committee was finalising a new standard. Without a statistical validation, the standard would have been published with no framework for its performance – no precision data, no measurement uncertainty. Is the new standard fit for purpose, and what can laboratories expect when they apply it?

The protocol.

A collaborative interlaboratory study under ISO/IEC 17043 and ISO 13528, designed around the draft standard itself and carried out on real samples. All committee members were involved in the study.

The outcome.

The standard was validated with the expected performance – measurement uncertainty included. Precision data nowaccompanies the standard, giving every future user a reference for what the method can deliver.

WHEN BILATERAL COMPARISONS ARE NOT ENOUGH – PORCELAIN VITREOUS ENAMEL

The scientific question.

An industry association, together with its national associations, faced an evolving legislative framework for its products. Before turning to the interlaboratory approach, considerable resources had been invested in engaging two accredited laboratories on a wide range of samples – and, as statisticians know well, nothing conclusive comes from bilateral comparisons: no inference, no interpretation.

The protocol.

A collaborative interlaboratory study under ISO/IEC 17043 and ISO 13528 on the release from objects covered with porcelain vitreous enamel, designed to characterise the test methods and the traceability of the elements under discussion, and to provide the sector with reference information on laboratory performance.

The outcome.

The association obtained the conclusions it needed to take an informed, data-based position on the new legislation for its products – conclusions that two years of bilateral testing had not been able to provide.

REAL MEASUREMENT UNCERTAINTY AGAINST A LIMIT UNDER DISCUSSION

The scientific question.

A legal limit was under discussion at European level. The organisation involved needed defendable data on one point that opinions could not settle: what is the real measurement uncertainty of the test method when applied across laboratories – and what does it imply for the limit under review?

The protocol.

A collaborative interlaboratory study under ISO/IEC 17043 and ISO 13528, with anonymous participation, designed to measure the real dispersion of results across laboratories and to derive the real measurement uncertainty of the method against the limit under discussion.

The outcome.

A complete statistical dossier, with interpretation, which was subsequently used in the European regulatory review process. The study is covered by confidentiality: sector and parties are not disclosed.

Conference presentations

The platform’s approach, presented to technical and scientific audiences. Slides are available for consultation.

• New regulations, new thresholds and one solution: holistic data validation for compliance – 1st episode. Annual technical conference, Baveno 2024. – English
• New regulations, new thresholds and one solution – 2nd episode / From regulatory risk to analytical reliability: using interlaboratory statistics to select competent laboratories. Annual technical conference, Baveno 2025. – English
• Proficiency Testing Collaborative Study “Interlaboratory in LCA” in the food and packaging sector. LCA FOOD 2024, international conference. – English
• Interlaboratory in LCA – results and the qualification of LCA practitioners. MAPPING LCA Annual Conference, Milan 2025. – English
• Incertezza nel calcolo di impatto (Uncertainty in impact calculaton). Advanced course for the Italian LCA Network, 2022. – Italian

Your study could be the next story told here – or the next one nobody will ever read about.

Public or confidential: you decide. Open a project, or talk to one of our specialists first.