| dc.description.abstract | This research addresses the challenges associated with conventional methods of performance-based building design, i.e., ignoring the existing uncertainties and lacking a systematic framework to incorporate risk assessment in building performance analysis. The goal of this research is to tackle data uncertainties and potential risks in architectural design decision-making with a focus on building energy performance.
A novel framework (BIMProbE) is created for integrating Building Information Modeling (BIM) into probabilistic building energy simulation to enhance the user interface and system interface for such simulations. In this research, BIM tools and BIM API are used to create probability distributions of material thermal properties for the building energy simulation. The present work enables a probabilistic BIM for energy simulation and future other building performance simulations. Also, BIM and parametric design tools (as the two major tools allowing a change of architectural design method) are used together for probabilistic design decision making.
The proposed framework is tested with three energy evaluation test cases. In each test case, building annual thermal load is measured for three different design options using deterministic and probabilistic methods. Also, three design decision making criteria including expected value, maximax, and maximin are applied to discuss the simulation results based on different attitudes towards risk. Different probabilistic distributions of input variables (normal or Poisson) are used in each test case. The thermal properties of building materials, building internal heat loads, HVAC system specifications, and some aspects of occupant behavior are considered as uncertainties to predict the probability distribution of building annual thermal load.
The results show that compared with the existing deterministic method for architectural design, using probabilistic methods is possible to result in significantly different design decisions to be made or different design options to be selected. Therefore, probabilistic methods should be considered in design simulation and decision-making. Furthermore, the extra information obtained from the probabilistic approach, including the mean, standard deviation, and variance, could help predict the possible range of the outcome for each design option. This research concludes that investigating probabilistic architectural design methods forms a major future research area in computational design. | en |
