Speaker
Description
This contribution presents an open-source Python workflow for variance-based global sensitivity analysis of upcycling life cycle assessment inventories, built on Brightway2 and the SALib library. The workflow enables practitioners to identify which foreground parameters drive result variability before committing to a full comparative study, addressing a common but under-supported step in LCA practice.
The workflow is demonstrated across six industrial upcycling cases from Danish SMEs, covering metal ceiling panels, concrete outdoor furniture, ceramic household products, acrylic serving trays, and paper-based office products. For each case, foreground inventories were constructed in Brightway and linked to ecoinvent. Sobol first-order and total-order sensitivity indices were computed across climate change and cumulative energy demand impact categories.
Results reveal both cross-case recurring sensitivity drivers and case-specific patterns tied to material system characteristics. The full workflow, including parameterised inventories, SALib configuration, and visualisation scripts, is shared as a reusable, openly licensed repository.
| How much time do you ideally wish for your contribution? | 15 min (Presentation, slides; Presentation, with notebook) |
|---|