Contribution to the development of a model for the analysis of stocks and flows of waste materials from buildings at a territorial scale

The building stock represents one of the pillars of our society. Nevertheless, its development, maintenance, and renewal requires a colossal amount of resources. In addition, these activities generate significant waste flows that will involve temporal and spatial impacts on the environment. Therefore, saving resources and better waste management are crucial challenges for the sector. In France, these challenges are tackled through recent anti-waste law for a circular economy and commitments for green growth. The first objective of this research work is to identify the multiple needs building stock portfolio managers have in order to achieve the Circular Economy objectives. This is explored through structured interviews with key stakeholders. The result is that one of the principal needs is to gain knowledge on the material composition of their building stock and the associated material flows. Such knowledge has to be detailed, reliable and must be deployed at several territorial scales in order to get data on the type, quantity, condition, age, and availability of the material stock. With such insight, the second objective of this research work is the development of a multi-scale material characterization of the building stock and the flows resulting from demolition. This characterization is done through a Material Flow Assessment (MFA) model called BTP-flux based on a component bottom-up approach that uses accessible national databases linked with GIS tools and structured data analysis techniques to make it reproducible to other territories. A case study on the Ile de France region (France) illustrates the model application. Estimations on 31 material categories are presented for different territorial scales. Finally, the robustness of the results presented by the model is tackled through the development of a specific methodology for calculating uncertainties from categorical variables. An adaptation of the confusion matrix and the Monte Carlo method is used to propagate uncertainty. The results allow to obtain confidence intervals for the material stock and waste flow estimation for the same Ile de France region case study. The developed methodology can be applied to any MFA model using categorical variables. The categorical variable analyzed in the case study is the materiality of the buildings' walls. The results show that the estimations of the stock and the waste flows, with the BTP-flux model, are robust up to the departmental scale due to the data availability and because, at that scale, both the random errors and the biases are relatively low. As a conclusion, these thesis results allow to support the deployment of reuse and recycling sectors by providing local authorities with robust and reliable multi scale information. Such information on the nature and quantities of materials is decisive for establishing strategies and action programs aimed at re-circulating resources at several territorial scales. In particular, this work allows the forward planning of predefining needs for storage, sorting, and waste recovery.