Proactive building simulations for early design support: Multi-actor decision-making based on Monte Carlo simulations and global sensitivity analysis

Simulations are commonly used to assess building performance with respect to energy demand and indoor environment. However, the use of performance simulations is limited during the early stages characterized by large uncertainties. This industrial Ph.D. study presents a novel simulation approach that relies on thousands of simulations representing the multidimensional design space. Interactive visualizations enable decision-makers to explore, in real-time, the vast design space and identify favorable solutions which satisfy the needs of different stakeholders. Sensitivity analysis helps reveal important design parameters that require the most attention when seeking to improve building performance. Fast metamodels facilitate immediate feedback on design changes and reduce time-consumption related to performance assessment. Ultimately, the work described in this thesis and on buildingdesign.moe.dk facilitates proactive guidance and supports collaboration between building owners, architects, engineers, and contractors. This helps the design team to create buildings with high performance and minimum costs