Data quality and uncertainty in LCI

Abstract

It has recently been acknowledged that the quality of data used in Life Cycle Assessment (LCA) is one of the most important limiting factors to the application of the methodology. Early approaches dealing with this problem solely based on Data Quality Indicators (DQI) have revealed their limitations, and stochastic models are increasingly proposed as an alternative. Although facing methodological and practical difficulties, for instance the characterization of the distribution of input data, these stochastic models can significantly enhance decision-making in LCA. Uncertainty and data quality, however, are two distinct attributes. No matter how sophisticated the stochastic models are, they do not address the issue of the adequacy of the data used with regard to the goal of the study. Actual data on the distribution of SO emissions for US coal fired power plants for instance, would be of low quality for a European study. It is therefore believed that mixed approaches DQI/stochastic models should be developed in the future.