Quantifying uncertainty in LCA-modelling of waste management systems

Abstract

26 27Uncertainty analysis in LCA studies has been subject to major progress over the last years. In the context of waste 28 management, various methods have been implemented but a systematic method for uncertainty analysis of waste-29 LCA studies is lacking. The objective of this paper is (1) to present the sources of uncertainty specifically inherent to 30 waste-LCA studies, (2) to select and apply several methods for uncertainty analysis and (3) to develop a general 31 framework for quantitative uncertainty assessment of LCA of waste management systems. The suggested method is a 32 sequence of four steps combining the selected methods: (Step 1) a sensitivity analysis evaluating the sensitivities of 33 the results with respect to the input uncertainties, (Step 2) an uncertainty propagation providing appropriate tools for 34 representing uncertainties and calculating the overall uncertainty of the model results, (Step 3) an uncertainty 35 contribution analysis quantifying the contribution of each parameter uncertainty to the final uncertainty and (Step 4) 36 as a new approach, a combined sensitivity analysis providing a visualization of the shift in the ranking of different 37 options due to variations of selected key parameters. This tiered approach optimizes the resources available to LCA 38 practitioners by only propagating the most influential uncertainties. Waste management has during the last decade been subject to a range of life cycle assessment (LCA; described in 48 ISO, 2006) studies e.g. Damgaard et al. (2011, Lazarevic et al. (2010) and Pires et al. 49 (2011). The purposes of these studies have been to help quantifying, for example, where in the waste management 50 system the environmental loads and savings are taking place, which technologies are preferable under specific 51 conditions, or the balance between material and energy recovery. LCA-models specifically focusing on waste 52 management systems are available; see Gentil et al. (2010) for a review of the models. 53As for any LCA study, results are subject to uncertainty due to the combined effects of data variability, 54 erroneous measurements, wrong estimations, unrepresentative or missing data and modelling assumptions. 55Uncertainty is of two different natures: while epistemic uncertainty relates to an incomplete state of knowledge 56 (Hoffman and Hammonds, 1994), stochastic uncertainty originates from the inherent variability of the natural world. 57Such uncertainty can be spatial (e.g. when the farming practice of land receiving compost varies spatially) or 58 temporal (e.g. when the performance of a process varies with time). These two different natures of the uncertainty are 59 usually treated together and referred to by the term "uncertainty". 60 They found that stochastic modelling was the most frequently-used method to propagate uncertainties in LCA. This 75 method propagates probability distributions using random sampling like the Monte Carlo analysis. However, they 76 noted that many of the studies using such modelling...

GHG emission factors for bioelectricity, biomethane, and bioethanol quantified for 24 biomass substrates with consequential life-cycle assessment

“…Sensitivity and uncertainty analyses were addressed at two levels: i) scenario uncertainties and ii) parameter uncertainties, conformingly with the approach suggested in Clavreul et al (2012). Scenario uncertainties were addressed by: I) assessing the performance of the bioenergy scenarios with two opposite energy systems (fossil/nonfossil).…”

Evaluating landfill aftercare strategies: A life cycle assessment approach

“…The SR is the ratio between the percentage change in the model's result and the percentage change in the parameter value (Clavreul et al, 2012), i.e. a parameter SR greater than 1 (as absolute value) implies that a variation of the parameter value induces a greater relative variation in the model output.…”

Contribution to the environmental impact assessment of biodiesel in the context of Spain

“…A distinction has been traditionally made between epistemic or systematic uncertainty, which is due to an incomplete state of knowledge, and stochastic or statistical uncertainty, which is observed every time the same experiment is run due to inherent variability of the natural phenomena. These two different natures of uncertainty are usually treated together and referred to by the term "uncertainty" (Clavreul et al, 2012).…”

“…Scenario analysis is presented as a tool for dealing with all these sources of uncertainty, thus keeping the what-if reasoning as much as possible out of the LCA methodology itself (Heijungs and Guinée, 2007). It consists of testing different options individually and observing the effect of these changes on the final result (Clavreul et al, 2012). This yields discrete results which relate the LCA outcomes with the corresponding assumption.…”

“…Obtaining the right uncertainty information and deciding which statistical distribution is appropriate is not an easy task (Heijungs and Frischknecht, 2005), and assumptions have to be made. As noticed by Clavreul et al (2012), probability distributions are often arbitrarily selected according to different criteria. The characterization of the most frequent probability distributions is, however, beyond the scope of the present dissertation.…”