Child dataset

A child dataset is a dataset resulting from combining a Parent dataset with a Delta dataset. This hierarchical structure, developed by ecoinvent, allows efficient management of LCA datasets with multiple variations whilst maintaining consistency and reducing redundancy.

Child datasets are based on the principle of data inheritance. A parent dataset is a template containing default information that describes a baseline activity. The delta dataset specifies only the differences needed to create a specific variant. When these two are combined, the resulting child dataset inherits all field values from the parent except where the delta dataset explicitly overrides or modifies them.

This approach offers several advantages for LCA database management. First, it significantly reduces redundancy by storing common information only once in the parent rather than repeating it across multiple similar datasets. Second, it maintains consistency across related datasets, as updates to shared characteristics in the parent automatically propagate to all child datasets. Third, it makes managing geographical, temporal, or technological variations more transparent and efficient, as only the actual differences need to be documented in the delta datasets.

For example, a parent dataset might describe a global average electricity production process, whilst child datasets would represent specific country or regional variations. The delta datasets would specify only the parameters that differ by region, such as the fuel mix, emission factors, or efficiency rates, whilst inheriting all other shared characteristics from the parent.

This hierarchical structure is especially useful in large LCA databases with numerous variants of similar activities, allowing database developers to maintain large collections of related datasets more efficiently and with fewer inconsistencies.