BIM geometric error detection for building energy model generation

The automatic generation of building energy models (BEMs) from Building Information Model (BIM) files has numerous challenges, including issues stemming from the poor quality of geometric data in the input BIM files and the geometric operations required to form the second-level space boundary (2LSB) surface set when not already provided by authoring tools. This study focuses on addressing the above issues in BIM data by examining and thoroughly analyzing the impact of geometric data quality of BIM input files during the transformation process. The geometric data errors found in BIM input files, which subsequently affect the quality of the resulting BEMs, are categorized into three distinct types: surface errors, clash errors, and space errors. Dedicated detection algorithms are introduced for each error type and applied to an IFC BIM model. The study identifies multiple geometric flaws that compromise BEM generation and demonstrates how targeted, algorithm-specific error detection substantially reduces the correction burden for energy modelers by focusing only on simulation-critical errors rather than all geometric inconsistencies. The novelty of this work lies in the formal mathematical characterization of error types using computational solid geometry, the development of dedicated detection algorithms based on these representations, and the demonstrated practical applicability through industry-standard export formats that support integration with BIM authoring tools.