Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
The Paris agreement on climate change called for carbon neutrality as of 2050. The built environment is one of the major contributors to the greenhouse effect, representing 39% of global emissions. As a result, this sector is targeted by green standards and regulations to set limits for building carbon emissions. Life Cycle Assessment (LCA) is widely recognized as an appropriate method to measure these emissions. This is particularly critical at the early design stage where the decisions that most influence the project are made. However, previous studies show that current LCA methods, both time-consuming and requiring a high resolution of detail, are most inadequate, which makes them too rarely used by practitioners today. This thesis aims to tackle this issue by proposing a novel approach to LCA adapted to the early design context. The first step was the identification of obstacles responsible for the current low use of LCA. This analysis was based on an extensive survey about the practice of 500 architects and engineers in Europe. Secondly, a literature review identified four appropriate techniques to overcome these obstacles: parametric assessment, sensitivity analysis, target cascading and data visualization. Combining them to the LCA led us to the data-driven method for low-carbon building design at early stages we adopted in this thesis, removing the need to set premature assumptions about future design developments. The method proposes a knowledge-database of design alternatives generated with a parametric approach that applies a combination of user-defined design options, using the Saltelli sampling technique, to a project-specific massing scheme. Later, the carbon emissions of each of these design alternatives are calculated. It is thus possible to explore thousands of alternatives and understand the consequences of architectural choices on the carbon emissions by using data visualization techniques. Moreover, the method proposes Sobol sensitivity indices quantifying the design parameter influence on the carbon emissions, in order to limit the scope of building components that designers should prioritize. Finally, the method specifies carbon budgets, the upper limit of carbon emissions a building component should not exceed, with the possibility to compare this budget with any product available on the market that would not have been included upfront as a design option in the parametric approach. To assess the usability of the method, a computer-based prototype was thereafter developed to get it tested in the frame of a real design project. By taking advantage of the architectural competition for the Smart Living Lab in Fribourg, Switzerland, this critical testing phase was carried out by asking the practitioners involved to actually use the developed prototype to meet the SIA2040 carbon objectives set for the project. This thesis proposes a new method for effectively integrating GHG emission targets into the early stages of the design process with the aim to increase both acceptance and use of LCA in the design field, but also to inspire developers towards a new generation of LCA software.
David Michael Volpe Ruggiero, Paola Costanza Miglietta