|Aspect||G-21 Background databases in LCA studies|
||According to the ILCD Handbook, the background system comprises those processes that are not under the direct control or decisive influence of the producer of the good. For attributional modelling these are typically processes for both the upstream and downstream parts of the entire life cycle of a product. For comparative assertions, the use of consistent background data is crucial. Generally, a product or a building process tree always includes intermediate flows (according to ISO 14044), such as energy, electricity, transport and raw materials. These processes may be modelled using background data available in generic databases.
In this context, which database should be used for background data in an LCA study? Which database should be chosen, or which rules should be defined for the use of a database for comparative assertions?
related study objective
|☒ stand-alone LCA||☒ comparative assertion|
related study phase
|goal and scope definition||inventory analysis (LCI)||impact assessment (LCIA)||interpretation||reporting|
|new buildings||existing buildings||building products||screening LCA||simplified LCA||complete LCA|
|Provisions||All data used in an LCA study should be modelled with a uniform methodology. This can typically be achieved by using a public or commercially available background database and modelling the foreground system in a consistent way.For comparative studies (in simplified or complete LCA studies), it is highly recommended, or even essential, to use datasets that are consistent, e.g. consistent within one database. If this is not possible (e.g. because of data gaps or incomplete documentation), any other data should comply with the methodological rules of the database used for the study.|
9 Life cycle interpretation
9.3.4 Consistency check
||In general, the LCA practitioner should use consistent data from a single source. For European research projects, the use of European databases such as ecoinvent, GaBi, ELCD or ESUCO may be appropriate. The mixing of background data from different databases should be avoided for comparative assertions. This is especially true if data from different databases are not used similarly in the compared models, or if the methodological rules and the quality guidelines do not match between the mixed data.However, in practice the mixing of background data can occur. For example:– if one background dataset (e.g. raw material production, or energy supply) is more representative for the context of the study than the one in the main database being used;– for EPD databases where no mandatory background database is provided. In this case the background data provider should ensure that the use of their data does not lead to bias in the comparative assertions (as the practitioner cannot influence the background data).In general, the use of different background data is not a problem; it is more the adaptation of data in terms of methodology and cut-off rules that can be a problem (for the LCA study). Some databases providing unit process data can easily be adapted to the goal and scope of the LCA study, and to the other background data that cannot be modified, but some databases cannot be adapted in this way.If the lack of specific datasets in a background database leads practitioners to combine data from different literature sources, they need to decide whether it is more important to use a consistent but roughly estimated dataset (and possibly data that are not relevant for the context of the study), or to use a dataset that may be better in terms of representativeness, but which is methodologically inconsistent with the alternative data.In this context, one possible solution is to use data quality indicators for assessing both the representativeness of the data and the consistency of the methodology, compared with the goal and scope of the study: for example, is the dataset compliant with the national context in terms of representativeness?|