Show simple document record

Other titre : Développement d'un modèle dynamique incorporant l'analyse de la variabilité et de l'incertitude pour l'évaluation des conséquences environnementales du cycle de vie des chaussées

dc.contributor.advisorBen Amor, Mourad
dc.contributor.advisorYahia, Ammar
dc.contributor.authorAzarijafari, Hessamfr
dc.date.accessioned2018-10-04T15:35:39Z
dc.date.available2018-10-04T15:35:39Z
dc.date.created2018fr
dc.date.issued2018-10-04
dc.identifier.urihttp://hdl.handle.net/11143/13618
dc.description.abstractThe selection of an environment-friendly alternative for pavements remains a challenge for governments for a long time now. A significant quantity of materials is required for the construction of pavements and maintenance of traffic service conditions above the performance thresholds. In order to meet the demand for construction materials, energy-intensive activities must be operated, which produce various types of emissions. Further, it should be mentioned that the environmental impacts of pavements are not limited to construction materials and machinery. In fact, pavement selection can induce changes in fuel consumption of vehicles and, to some extent, can adjust the heating and cooling demands of buildings through the properties of pavements. So far, several environmental impacts from alternative materials for constructing pavements have been investigated through the method of life cycle assessment (LCA). Conducting LCA of pavement on a large scale, such as, when considering its policy-related implications or requirements, is distinguished from small scale assessment. In the large-scale assessment, it is important to assess the changes induced through the demand of other products with the same function and/or the induced production volume of the co-producing processes. Consequential LCA (CLCA) is employed to capture the direct and indirect environmental consequences of pavement selection in different sectors. Previous CLCA studies have only captured the long-term affected technologies in distinct phases of the product life cycle. Hence, the results tend to neglect the short-term impacts and the impacts of short-run product systems. In fact, the lack of dynamic accounting in CLCA prevents policymakers from gaining a more informed and comprehensive analysis of emission flows over time. Furthermore, given the long use phase of pavement infrastructures, the vast number of interdependent parameters constantly change as a function of time. A dynamic structure is essential to link these parameters to dynamic characterization factors (CFs). Moreover, apart from being complex and requiring the prediction of technological improvements, applying the emission factors and fuel consumption or efficiency improvements can help enhancing the accuracy of the decision-making process involved in the selection of pavements. However, the temporal distribution of impacts induced by pavement life cycle parameters is not fully captured in previous CLCA studies. Moreover, the variations of the results in CLCA studies are usually overlooked and are not included in the interpretation of the conclusions. The uncertainty and variability sources should be investigated to examine the robustness of the results. To reach a comprehensive model for the uncertainty analysis, it is essential to consider the interdependencies of the variability and uncertainty sources. In fact, neglecting the interdependencies may lead to an entirely different conclusion compared to those obtained through an independent sampling. So far, no method has been proposed to consistently assess the variability and uncertainty sources of pavements LCAs either in consequential or attributional frameworks. To fill in the research gaps, this dissertation aimed to develop a comprehensive framework, that is able to capture interdependencies of parameters while analyzing uncertainty and variability sources. The proposed model comprises Monte Carlo simulations to propagate the sources of variations to the results. The model was applied to a case study of attributional life cycle assessment (ALCA), in which asphalt and concrete pavements were compared. Different sources, such as uncertainty due to data quality and methodological choices as well as variability of parameters, were investigated. The results of the Monte Carlo analysis show that it is feasible to assess the combined and the individual effects of common uncertainty and variability sources. Based on the variability and uncertainty of the results, a certain conclusion is case specific at both the midpoint and endpoint levels was identified. The methodological choices, such as allocation, can change the environmentally preferred scenario in four midpoint categories. The variability in construction materials and methods can change the preferred scenario in different damage categories, such as, human health and global warming. In addition, the preferred scenario in ecosystem quality can be changed when the parameter uncertainty is taken into account. The reason is that the worst qualitative scores are given to the geographical uncertainty of the elementary flow that majorly contributes to ecosystem quality (i.e. zinc). The combined effect of the uncertainty and variability sources for this case study prevents the decision-maker from reaching a robust conclusion about the ecosystem quality, human health, and global warming effects. However, for the resources category, the comparative result between asphalt and concrete pavements is sufficiently large. Therefore, the sources of variability and uncertainty do not change the preferred scenario. The second objective of this research project is to develop a dynamic consequential framework for assessing the environmental aspect of pavements. Moreover, the proposed model in the first objective was adopted to assess various sources of uncertainty and variability, such as data quality and modeling and variability, in different life cycle phases of pavements. The dynamic changes in the demand vector and the technosphere matrix were computed considering the time horizon of the affected technologies. The life cycle components were delineated through a precise parametrization of more than 130 time-dependent factors. According to the uncertainty analysis model proposed in the previous stage, a Monte Carlo simulation was conducted to propagate the variations in the system dividing the parameters to data quality uncertainty, modeling uncertainty, and variability. The proposed method was applied to the case study of shifting from an asphalt pavement (business-as-usual) to a concrete one (alterative). The obtained results show that a simplification for capturing the dynamic changes can result in an inaccurate assessment of the damage results. The environmental benefits credited by substituting the business-as-usual scenario with the alternative is overestimated by 7, 17, and 77% for climate change, ecosystem quality and resources categories, respectively. In addition, the results of ecosystem quality category show that including the dynamic spirit in the modeling can result in a 114% higher impact than implementing the static approach. The lack of accounting for temporal profile of the greenhouse gas (GHG) emissions in static CF results leads to an overestimation of the global warming potential (GWP) benefits of substituting asphalt with concrete by 473.5 t CO2eq. The uncertainty results show 41-71% contribution of the variability sources to the variance of the four damage categories. This variability is mainly attributed to the monthly temperature accounting (8-16% to the variance of damage results) and the service life of pavements (11-15% to the variance of damage results). Despite propagating the various sources of uncertainty and variability in this case study, the conclusion on the shifting decision remained unchanged in the damage results. This research project clearly showed the impact of considering the dynamic changes in the environmental assessment of pavement. More specifically, when it comes to the use phase and the time horizons of the affected technology, it is essential to include the time-dependent variables to improve the representativeness of the obtained results. Nevertheless, the importance of uncertainty and variability sources and distinguishing them in the pavement LCA should not be ignored.fr
dc.language.isofrefr
dc.language.isoengfr
dc.publisherUniversité de Sherbrookefr
dc.rights© Hessam Azarijafarifr
dc.subjectPavementsfr
dc.subjectConsequential life cycle assessment (CLCA)fr
dc.subjectUncertainty analysisfr
dc.subjectAttributional LCAfr
dc.subjectDynamic inventoryfr
dc.subjectIndirect effectsfr
dc.subjectEnvironmental policy-makingfr
dc.titleDevelopment of a consequential dynamic model incorporating variability and uncertainty analysis for assessing life cycle environmental impacts of pavementsfr
dc.title.alternativeDéveloppement d'un modèle dynamique incorporant l'analyse de la variabilité et de l'incertitude pour l'évaluation des conséquences environnementales du cycle de vie des chausséesfr
dc.typeThèsefr
tme.degree.disciplineGénie civilfr
tme.degree.grantorFaculté de géniefr
tme.degree.levelDoctoratfr
tme.degree.namePh.D.fr


Files in this document

Thumbnail

This document appears in the following Collection(s)

Show simple document record