Paper Conference

ASHRAE & IBPSA-USA SimBuild 2016: Building Performance Modeling Conference


Regression-Based Building Energy Performance Assessment Using Building Information Model (BIM)

Mohammad Asl, Weili Xu, Jin Shang, Barry Tsai, Ian Molloy
Autodesk Inc., San Francisco CA
Carnegie Mellon University, Pittsburgh, PA

Abstract: The building sector, the largest consumer of the United States primary energy, has been seeking to take necessary actions to reduce energy use of buildings while keeping them functional and comfortable. As a result, building designers are increasingly expected to improve the energy performance and reduce the carbon footprint of their building models in the design process. Higher energy savings opportunities are available during early stages of design; however, energy performance assessments are typically done during the later phases due to the lack of the easy-to-use and efficient tools that could help architects explore design alternatives. In this paper, we introduce a new Formbased Energy Performance Regression Model (FEPRM) that is capable of real time quantification of the building energy performance based on the form variation in a building information modeling (BIM) tool. FEPRM generates a set of models and simulations based on the project specifiations and provides a seamless energy performance feedback to the designer for decision making during the conceptual design process using a regression-based model. In this paper, we investigated the use of FEPRM on 3 different cases with different AHSRAE climate zones to demonstrate the usefulness of the proposed method. The comparison between actual simulation and predicted energy performance is provided to study the accuracy of the FEPRM; moreover, a set of extended test cases are also described.
Pages: 357 - 364