A critical perspective on uncertainty appraisal and sensitivity analysis in life cycle assessment

Abstract

In this study, we review approaches for uncertainty appraisal in the life cycle assessment literature. We cover the acknowledgment of stochastic and epistemic uncertainty in uncertainty and sensitivity analysis and knowledge quality assessment, respectively. Consistent with previous works, our findings indicate that uncertainty is only appraised in a few studies on life cycle assessment. Most of these contributions cover only one of the phases of life cycle assessment, mainly the life cycle inventory phase. Less attention has been devoted to the phases of goal and scope definition and life cycle impact assessment. Additionally, in most studies, uncertainty analysis and sensitivity analysis have been applied independently, as it is wrongly assumed they cover different uncertainty spaces. We also identify the scope for improvement in the appraisal of epistemic uncertainty and the correct definition of the probability distribution of the uncertain factors. We conclude by highlighting studies in which sensible practices have been adopted, identifying open challenges, and suggesting possible ways forward.

Model uncertainty versus variability in the life cycle assessment of commercial fisheries

“…High sensitivity to assumptions and high uncertainty represent recognized problems with models used for decision‐making in the sustainability domain (Saltelli et al., 2020). Within the context of LCA, it is well‐known that modeling choices can substantially affect the results (Lo Piano & Benini, 2022), particularly when using different approaches to the solving of multifunctionality and the definition of supply mixes. The influence of modeling choices on results has been explored in several studies across different sectors and products such as wood (De Rosa et al., 2018), meat (Wilfart et al., 2021), biorefinery products (Sandin et al., 2015), bioenergy (Brandao et al., 2022; Wardenaar et al., 2012), and fish (Avadí & Fréon, 2013), just to mention a few.…”

“…While it is arguably difficult to rigorously classify uncertainty in the context of LCA, various recent studies have attempted this (Brandao et al., 2022; Clavreul et al., 2012; Igos et al., 2019; Lo Piano & Benini, 2022). Generalizing across them, it is possible to divide uncertainty between two main distinct sources: epistemic and aleatory .…”

Relative sensitivity value (RSV): A metric for measuring input parameter influence in life cycle assessment modeling

“…Most attributional LCA tools utilize a process matrixbased approach to quantify the inventory and impacts associated with a product (Igos et al, 2019). A typical attributional LCA model, as shown in Figure 1, can be treated as a function, f, which accepts input parameter arguments, X x x x x , , …, , …, Currently, only a small fraction of peer-reviewed LCA case studies report on uncertainty assessment and sensitivity analysis (Bamber et al, 2020;Lo Piano & Benini, 2022). The topic of uncertainty in impacts due to epistemic uncertainty in input parameters is gaining ground in LCA-related research, as LCA practitioners begin to recognize its importance in accurate decision-making (Groen et al, 2017;Lo Piano & Benini, 2022;Scrucca et al, 2020;Wei et al, 2015).…”

“…A typical attributional LCA model, as shown in Figure 1, can be treated as a function, f, which accepts input parameter arguments, X x x x x , , …, , …, Currently, only a small fraction of peer-reviewed LCA case studies report on uncertainty assessment and sensitivity analysis (Bamber et al, 2020;Lo Piano & Benini, 2022). The topic of uncertainty in impacts due to epistemic uncertainty in input parameters is gaining ground in LCA-related research, as LCA practitioners begin to recognize its importance in accurate decision-making (Groen et al, 2017;Lo Piano & Benini, 2022;Scrucca et al, 2020;Wei et al, 2015). Scenario analysis is the most commonly used method to estimate uncertainty, in which impacts from two or more distinct sets of input parameters are compared to provide a range of possible impact scores (Gawron et al, 2018;Wang et al, 2018;Welz et al, 2011).…”

“…Currently, only a small fraction of peer‐reviewed LCA case studies report on uncertainty assessment and sensitivity analysis (Bamber et al, 2020; Lo Piano & Benini, 2022). The topic of uncertainty in impacts due to epistemic uncertainty in input parameters is gaining ground in LCA‐related research, as LCA practitioners begin to recognize its importance in accurate decision‐making (Groen et al, 2017; Lo Piano & Benini, 2022; Scrucca et al, 2020; Wei et al, 2015).…”