How the Use of Different Databases Affects the Comparability of Life Cycle Assessment Results of Fibre Reinforced Polymer Composites

Kurzfassung

Sustainability gained significant popularity these days and in order to support decision making towards this, Life Cycle Assessment (LCA) is becoming increasingly important. However, the full potential of these assessments can only be unlocked if the comparability of the results is guaranteed. Databases such as Ecoinvent and Sphera are the backbone of LCA models and they have a crucial impact on the results. Usually, foreground data for a product or service is limited to a small range of known activities in a gate-to-gate perspective where primary data can be obtained. For a more complete view on the potential environmental impacts, it is necessary to extend the model with the upstream activities back to the extraction of natural resources to gain a cradle-to-gate or a full life cycle perspective (cradle-to-grave/cradle). When talking about the comparability of results, it is important to address the issue of uncertainties. These exist in every phase of LCA and are too many to point out individually here. The existence of different databases, intended to provide support in LCA, are just a part of this bigger picture. It is therefore not only important to know which databases are utilized for analysis in order to be aware of their differences and resulting uncertainties in this particular context, but also to understand their differences in more detail. One important question is if it is valid to mix different databases to fill gaps in the Life Cycle Inventory. This study intends to make potential differences of databases more tangible. To achieve this the model for producing a typical fiber reinforced polymer (FRP) product made out of glass fibers and epoxy resin is created based on DLRs expertise in the manufacturing of lightweight systems. The foreground data for the selected product is fed with background data from different existing databases. The results obtained are finally used to unveil the differences for the sixteen categories of the EF3.1 methodology and to identify to what extent the comparability is affected. Additionally, the study gives an overview of the main uncertainty contributors in this use case as a contribution towards a transparent decision making.