ADVANCING THE INTEGRATION OF PERSISTENT AND MOBILE SUBSTANCES IN LIFE CYCLE IMPACT ASSESSMENT

Licentiatavhandling, 2024

Among the many chemicals we use in daily life, 'forever chemicals' such as per- and polyfluoroalkyl substances (PFAS), can persist and accumulate in the environment and may cause long-term harm. These substances can remain with us for an extended period not only in the products we use but also as contaminants in the environment. They can also slowly transform to form toxic byproducts.

There is an increased awareness and concern about the use of these chemicals, leading to efforts at both the regulatory level and within academic, industrial, and societal spheres. Risk assessments for chemicals have typically focused on either emission during use or from manufacturing facilities rather than the whole life cycle. On the other hand, life cycle assessment (LCA) often considers greenhouse gas emissions but omits toxicity concerns. There is a need for tools that can quantify the impacts of these chemicals along product life cycles, integrating them into LCA. Without calculating toxicity characterization factors (CFs) using tools such as USEtox, LCA is unable to quantify ecotoxicity impacts of chemicals. In addition, there are limitations regarding the availability of ecotoxicity data required to calculate these CFs. Even when data is available, there are concerns about the applicability of existing tools to accurately calculate CFs.

The research presented in this licentiate thesis is based on two studies. The aims are to determine the gaps in the availability of ecotoxicity CFs for persistent and mobile substances, examine the influence of ecotoxicity data selection and harmonization alternatives on Effect factors (EFs), and explore the calculation of extrapolation factors to convert effect concentration indicators. The first study offers methodological contributions on the harmonization of ecotoxicity data of chemicals for use in calculating CFs. It also assesses whether QSAR-based data can effectively replace experimental data. Additionally, it offers practical use values for ecotoxicity CFs of persistent and mobile chemicals, which were previously unavailable in existing USEtox database, thus supporting their inclusion in LCA studies. The second study addresses the uncertainty in the extrapolation factors when converting different effect concentration indicators (endpoints). This aids in reducing the uncertainty of using generic extrapolation values for chemicals by providing species group-specific extrapolation values for EC10 and EC50 effect concentration indicators.

The study concludes that the coverage of persistent and mobile chemicals in the USEtox database (version 2.01) is only 28% for the chemicals in focus in this thesis (18 out of 64). Consequently, emissions of chemicals lacking CFs cannot be included in ecotoxicity impact assessments in LCA due to the absence of CFs. Additionally, the ecotoxicity data harmonization approach can significantly influence the calculation of EFs. A pragmatic harmonization approach is recommended to ensure the process is feasible without compromising the accuracy and reliability of the harmonized data. Furthermore, QSAR methods should be considered a last resort when experimental ecotoxicity values are unavailable. QSAR methods lack accuracy in estimating ecotoxicity values for EF calculations. Finally, extrapolation factors at the species group level differ considerably from those at the generic level, leading to the conclusion that species-level factors should be used to reduce uncertainty in the extrapolated effect concentration indicators.

Several challenges have also been identified which should be addressed for LCA to contribute to the quantification of impacts of forever chemicals. These include adaptation of the USEtox model to persistent and mobile chemicals, and the availability of ecotoxicity data for these chemicals.