Keywords
life-cycle assessment; sensitivity analysis; uncertainty analysis; pedigree scores; knowledge-quality assessment
Start Date
15-9-2020 9:00 AM
End Date
15-9-2020 9:20 AM
Abstract
Life-cycle assessment (LCA) is a methodology developed for assessing environmental impacts associated with production and consumption of goods and services. Its use is upheld by international standards such as ISO14040 and ISO14044. On the top of this, prominent international policy makers - such as the European Union – recommend its use in policy impact assessments. LCA rests on four pre-analytical phases: i) goal and scope (definition); ii) inventory analysis; iii) impact assessment; and iv) (results) interpretation. Each one of them encompasses several stages that require methodological assumptions and choices about the parameters and processes inquired. In this contribution, we critically review the practices put in place for uncertainty appraisal and propagation in LCA by reviewing the relevant literature. Our analysis discusses both the epistemic – the so-called unknown unknowns and thus unquantifiable – uncertainty and the stochastic – the so-called known unknowns and thus quantifiable – parts of uncertainty. Besides the characterisation of uncertainty, we also examine how uncertainty is apportioned onto the input parameters and assumptions through sensitivity analysis (SA). While tools for the characterisation of epistemic uncertainty do exist and are used in LCA, the way in which these are employed is wanting. The same applies to knowledge quality assessment techniques, particularly for what concerns sample parameters distributions. Analogously, the same criticism can be applied to how the input space is sampled: uncertainty is typically propagated over only one, or a limited set, of the aforementioned analytical stages. As regards SA, one-variable-at-a-time approaches are largely prevailing in the literature. LCA practitioners could benefit from closer collaborations with experts in knowledge quality assessments and SA. The robustness of LCA findings could in this way be strengthened, while lights could be shed on both limitations and applicability of LCA to address policy-relevant questions in the context of uncertainty.
Uncertainty appraisal in life-cycle assessment
Life-cycle assessment (LCA) is a methodology developed for assessing environmental impacts associated with production and consumption of goods and services. Its use is upheld by international standards such as ISO14040 and ISO14044. On the top of this, prominent international policy makers - such as the European Union – recommend its use in policy impact assessments. LCA rests on four pre-analytical phases: i) goal and scope (definition); ii) inventory analysis; iii) impact assessment; and iv) (results) interpretation. Each one of them encompasses several stages that require methodological assumptions and choices about the parameters and processes inquired. In this contribution, we critically review the practices put in place for uncertainty appraisal and propagation in LCA by reviewing the relevant literature. Our analysis discusses both the epistemic – the so-called unknown unknowns and thus unquantifiable – uncertainty and the stochastic – the so-called known unknowns and thus quantifiable – parts of uncertainty. Besides the characterisation of uncertainty, we also examine how uncertainty is apportioned onto the input parameters and assumptions through sensitivity analysis (SA). While tools for the characterisation of epistemic uncertainty do exist and are used in LCA, the way in which these are employed is wanting. The same applies to knowledge quality assessment techniques, particularly for what concerns sample parameters distributions. Analogously, the same criticism can be applied to how the input space is sampled: uncertainty is typically propagated over only one, or a limited set, of the aforementioned analytical stages. As regards SA, one-variable-at-a-time approaches are largely prevailing in the literature. LCA practitioners could benefit from closer collaborations with experts in knowledge quality assessments and SA. The robustness of LCA findings could in this way be strengthened, while lights could be shed on both limitations and applicability of LCA to address policy-relevant questions in the context of uncertainty.
Stream and Session
false