Variability, Prospectivity, and Multifunctionality in Life Cycle Assessment
Year: 2025
Type: dissertation
Abstract: In this dissertation, uncertainties in life cycle assessment (LCA) are investigated. LCA is a method for the quantitative assessment of the potential environmental impacts of products, services, production processes, and companies over their entire life cycle. Since an LCA model is a simplified representation of reality, uncertainty exists regarding how accurately the assessed potential environmental impacts reflect the actual effects on environment. A source of this uncertainty is the inherent variability of input data. In many LCAs, real-world variability is not represented, as input data is often aggregated into average values for modeling. For instance, data on the composition of the fluctuating electricity mix in the course of a year is typically aggregated to a single yearly average. This uncertainty can be examined and reduced by using high-resolution data that captures the variability. However, the consideration of non-aggregated data poses a challenge for data collection, as well as for the calculation and interpretation of LCA results. Another source of uncertainty in LCA results is the methodological handling of processes that fulfill more than one function. An example of a so-called multifunctional process is pig iron production, which concurrently generates the coproduct blast furnace slag that can be used directly in cement pro-duction. As a result, environmental impacts occur simultaneously for several coproducts, which must be virtually assigned to the individual coproducts in the model. However, the necessary methodological choices for this assignment cannot be unambiguously derived from the process itself and therefore lead to uncertainties or ambiguities in the LCA. Furthermore, results are subject to additional uncertainty when the potential future environmental impacts of a process are assessed using a prospective LCA (pLCA). This approach involves the formulation of assumptions concerning, for instance, technological developments, boundary conditions, and scaling effects. This may include, among other aspects, the extrapolation of laboratory-scale data to an industrial scale or projections of the future electricity mix. The associated uncertainties arise in particular from limited data availability and quality, as well as the necessity of numerous assumptions and methodological choices. To address these uncertainties, scenario analysis is frequently applied in pLCA to represent different possible development pathways. Against this background, this dissertation investigates and develops approaches for addressing uncertainties that stem from data variability, multifunctionality, and future developments in LCAs using four case studies. In case studies on charging profiles of battery electric vehicles (BEV) and steel production, the LCA models were constructed modularly to efficiently incorporate high-resolution input data. In the BEV case study, a mixed-integer linear optimization model was also implemented. Using these approaches, LCAs were conducted for the first time for both ecologically optimized charging profiles and all steel grades produced at a steel mill. In further case studies on CO₂ storage in minerals and the recovery of secondary resources from waste incineration bottom slag, the influence of different approaches for handling multifunctionality on LCA results is analyzed. In addition, scenario analysis is used to evaluate the influence of future developments on the results of these case studies. The results of the case studies on BEV charging profiles and steel production show that considering data variability throughout the four phases of LCA substantially improves the informative value and relevance for decision support of LCA results. Moreover, it is emphasized that methods from operations research can support the analysis and interpretation of the results. The four case studies illustrate that pLCAs of multifunctional processes require the consideration of additional challenges compared to conventional LCAs. In scenario analyses, different assumptions regarding the handling of multifunctional processes can lead to substantial differences in the assessment of potential environmental impacts. To systematically address these uncertainties, a new methodological approach is developed in this dissertation, which combines and complements existing methods for handling multifunctionality with scenario analysis. This approach facilitates the analysis of factors that influence the identification and handling of multifunctional processes, as well as their consideration in pLCAs. In summary, this dissertation contributes to the further development of LCA methodology by investigating and developing approaches for the consideration of variability and multifunctionality in both conventional and prospective applications.
Author: Zacharopoulos, Leon
Language: English
Subfield: Environmental Engineering
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7106342989
Cites:
Cited by:
Related to: