Are ABMs the right framework to model the environmental impacts of the construction and demolition sector? Technical and methodological challenges of modelling
Résumé
Reusing construction materials offers substantial potential for reducing the environmental impact of the construction and demolition sector. However, implementing circularity strategies requires a spatially detailed understanding of material stock dynamics, which presents both methodological and computational challenges as the level of detail increases. This study introduces a stock-driven, rule-based simulation that forecasts – over space and time – material stocks, flows and associated embodied carbon from construction, renovation, and demolition activities. Currently, each agent in the model represents a building, allowing exploration of how factors such as recycled content, demolition and renovation rates, and types of new construction affect material flows and embodied carbon. However, as the model scales towards component or product-level detail, where individual building components become agents, the computational load increases significantly, raising questions about the feasibility of agent-based models at this level of granularity. Using Gothenburg’s residential building stock, we demonstrate the simulation’s capabilities for tracking eight building materials (and their embodied carbon) across two scenarios: a baseline and an energy efficiency policy. We discuss the current model’s limitations when increasing detail and the methodological challenges ahead in balancing model complexity with computational efficiency, particularly in addressing whether agent-based models remain suitable as the level of detail escalates from buildings to individual components.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |