Building Energy Modeling (BEM) plays a critical role in reducing energy con-sumption in the Architecture, Engineering, and Construction (AEC) industry. How-ever, creating accurate BEM models demands substantial expertise and effort. Build-ing Information Modeling (BIM) offers an opportunity to automate this process by converting BIM data into BEM models, but this approach faces two main challenges: ensuring accuracy in BIM-based BEM models and managing high simulation times and computational loads, especially for large projects. This thesis presents a methodology using BIM space boundaries to create thermal zones, analysed with various thermal zoning strategies and compared against a detailed base scenario. Results indicate that BIM-based BEM models show accuracy comparable to manually generated models. The zoning strategies lead to significant simulation time reductions of 82% to 90% and reduced energy load predictions by 10% to 20% when zoning is maintained on the same floor. Sce-narios merging zones across floors result in further energy load reductions due to Gross Floor Area (GFA) changes, though they may affect temperature uniformity. In conclusion, the methodology offers a balance between faster simulation times and accurate energy predictions, supporting informed decision-making in de-sign. Future research could integrate large language models (LLMs) to improve space identification and enhance thermal zoning automation.     «

Building Energy Modeling (BEM) plays a critical role in reducing energy con-sumption in the Architecture, Engineering, and Construction (AEC) industry. How-ever, creating accurate BEM models demands substantial expertise and effort. Build-ing Information Modeling (BIM) offers an opportunity to automate this process by converting BIM data into BEM models, but this approach faces two main challenges: ensuring accuracy in BIM-based BEM models and managing high simulation times and computational loa...     »