Life cycle inventories of electricity supply through the lens of data quality: exploring challenges and opportunities

Electricity is one of the main contributors to global environmental impacts, and its role as an energy carrier is expected to grow substantially. Consequently, reliable and accurate inventories of material and energy flows associated with electricity supply are essential in environmental assessments. This article aims to summarize existing challenges and opportunities in the modeling of life cycle inventories (LCIs) of electricity supply from a data quality perspective.

We systematically review the state-of-the-art in LCI modeling of current and future electricity supply worldwide. The analysis is structured according to the data quality characteristics proposed in ISO 14044: 2006: representativeness, completeness, consistency, reproducibility, uncertainty, data sources, and precision.

Looking at existing LCIs through the lens of data quality, we observe difficulties in obtaining temporally and technologically representative data, while geographically representative data is still unavailable for some regions. Moreover, meta-analyses encountered issues of reproducibility combined with a lack of consistency across studies, impeding interstudy comparability. Additionally, some flows such as upstream fugitive emissions have been underestimated. The aforementioned issues have a negative impact on the quality of LCIs. Here, we provide recommendations on how several methods such as equilibrium models, regression, or parameterization can be used to improve data quality, underpinned by more powerful data formats. Open-source models, data platforms, as well as a list of key parameters to be reported are suggested to facilitate reproducibility and enhance transparency of electricity LCIs.

There are several methods and resources that can be used to improve LCIs of electricity supply, enabling more ambitious and powerful analyses. Nonetheless, special care should be taken concerning tradeoffs between different data quality aspects. For instance, more complex and accurate models may result in a loss of transparency and reproducibility unless additional reporting efforts are conducted. Other approaches, such as systematic parameterization do not compromise data quality and should be used to improve the consistency and reproducibility of inventories.