Addressing uncertain scoring and weighting factors in social life cycle assessment

Abstract

Purpose

Social life cycle assessment (SLCA) is an approach to assess social performances over the entire product’s life cycle. Type I SLCA consists in a qualitative analysis of product’s activities through a set of subcategory indicators. To facilitate decision-making, it is useful to aggregate the subcategory indicator performances of the different activities along the product system into stakeholder dimensions. Currently, these aggregation steps are based on value judgments as scoring and weighting factors. We identified two types of uncertainties related to these value choices: (i) the uncertainty related to the scoring choice and (ii) the uncertainty related to the weighting factors. However, no current SLCA research considers the uncertainty of these value choices. The main objective of this work is to propose a method to systematically address uncertain scoring and weighting factors in SLCA.

Methods

Our method is structured in three phases: (i) assessing the uncertainty of the scoring factors, (ii) assessing the uncertainty of the subcategory indicators aggregation into stakeholder dimensions, and (iii) interpreting stochastic SLCA results. Monte Carlo simulations were conducted in order to generate stochastic SLCA results from randomly chosen scoring and weighting factors from value judgments of SLCA experts. We applied the approach on an illustrative case study comparing different options of biodiesel supply.

Results and discussion

Stochastic social impact scores were obtained for three considered stakeholder dimensions by propagating the uncertainty of the scoring and the weighting factors. Interpretation of stochastic SLCA score was performed by analyzing the first rank probability of each supply option within each stakeholder dimension. Results show a stable first rank position (100%) of one among the four options in two stakeholders’ dimensions (workers and society). For the local community stakeholder dimension, three out of the four supply options were likely to take first rank with a likelihood of 91, 8 and 1%, respectively.

Conclusions

Addressing uncertainty in social life cycle assessment allows to explicitly include the unavoidable variability of value choices adopted to aggregate the social performance evaluation of each activity over the product life cycle and across subcategory indicators into few meaningful stakeholder dimensions indicators. We demonstrated that our stochastic approach is key to inform decision makers about the robustness of SLCA results and showed that the ranking of a deterministic approach can be potentially modified when uncertainty is included.