Automated building geometry transformation and its validation from IFC BIM to BEM

Abstract

Building energy modeling (BEM) is a powerful and computerized approach for assessing building energy performance, which has been an increasingly common and important factor considered in building design and operation decisions. Building geometry is an essential input to BEM, and its definition generally consumes the largest portion of the effort in BEM input preparation. Building Information Modeling (BIM) contains precise building geometry information that can be utilized for BEM. Previous studies have shown that automating BIM-to-BEM geometric transformation can achieve significant time savings, error reduction, and inappropriate human interpretation elimination over traditional manual BEM creation. However, these studies have limitations that make it challenging to handle large-scale BIMs with complex building geometries. The main limitations include: (1) requiring considerable manual efforts to prepare high-quality input BIMs; (2) lacking the curved geometry transformation capacity; and (3) having room for efficiency improvement to process large-scale and complex BIMs. Furthermore, to ensure a reliable output of BEM geometry, the automatic transformation process needs to be rigorously validated. However, there still lacks reliable and efficient approaches/tools. This research addresses these limitations by proposing a novel Industry Foundation Classes (IFC) BIM-to-BEM geometric transformation (IBGT) approach and an automatic rule-based IBGT validation system (RIVS). The proposed IBGT approach consists of three successive steps. Firstly, by using a footprint-based faceting algorithm, an input IFC model is pre-processed by automatically detecting curved building objects and faceting their geometries into planar face-based polyhedrons. Secondly, a highly-efficient and error-tolerant approach is developed to enrich the pre-processed IFC model with 2nd-level SBs, the exact geometric data required by BEM. By applying multiple geometric representations (i.e., triangulated polyhedron, polyhedron, and axis-aligned bounding box) of building objects, the proposed SB enrichment approach can efficiently process large-scale IFC models. Furthermore, it tolerates certain common geometric errors in IFC models and can distinguish essential building objects of interest, which enable the IBGT approach to effectively reduce manual remodeling efforts for preparing the input IFC model. Thirdly, a rule-based schema-level conversion system is developed for converting the enriched IFC model into an EnergyPlus IDF model. The proposed RTVS provides an automatic, reliable, and user-friendly means to validate the transformed BEM geometry (i.e., 2nd-level SBs) stored in the enriched IFC model. It hardcodes a total of 44 rules to systematically validate 2nd-level SBs from the aspects of syntactic and semantic correctness, geometric correctness, and consistency. The proposed IBGT approach and RTVS are validated with a series of real-world IFC models separately as well as a case study using a complex real-life building holistically. The results indicate that by integrating the IBGT approach and RIVS, practitioners can efficiently obtain high-quality BEM geometry from large-scale IFC BIMs that contain curved geometries and certain common quality issues. This research thus contributes to the body of knowledge by developing an error-tolerant, curved geometry compatible, and highly-efficient IBGT approach as well as an automatic IBGT validation system. Both provide a solid foundation for researchers and industry software vendors to achieve the ultimate goal of complete information exchange from BIM to BEM.