Industry Foundation Classes (IFC) and Industry Foundation Classes XML (IFCXML) are widely accepted data frameworks and formats for information storage and exchange among building information models (BIM). IFCXML are comprehensive and compatible with the major BIM platforms; however, many studies suggest their complex structure often results in redundancy and inflexibility. It makes IFC and IFCXML unable to be applied in many practical application scenarios. This research first optimizes IFCXML, and proposes a series of automatic deduplication algorithms to solve its large amount of repeated information. Afterwards, the optimized IFCXML is applied to the building group energy simulation and building life cycle assessment, giving IFC series documents a broader application prospect.

For IFCXML optimization, researchers have proposed various IFC compression methods to reduce file size and restructure data organization, such as partial model extraction, Solibri IFC Optimizer, and ACC4IFC. However, simplification often results in missing information or data leakage, and there has never been a compressor specific to IFCXML. To overcome these issues, this study proposed a conservative compression method that removes duplicated information while maintaining the data structure of IFCXML through an iterative reference mapping method. Based on the data structure and geometry of IFCXML, the algorithm identified three kinds of duplicate information: independent entity duplication (IED), cross-reference entity duplication (CED), and property-set entity duplication (PED). To validate the proposed compressing method, this study conducted a validation test on six typical BIM models and benchmarked with other existing compressors (Solibri IFC Optimizer and ACC4IFC). The outcomes suggested that the proposed model could compress IFCXML files by 18%-59% without losing information.

For building group energy simulation, CityGML proposed different levels of details (LOD) for buildings, and LOD2 is the most recognized level for various energy analyses. However, building energy simulation requires detailed information on the geometry and thermal properties, LOD2 is not sufficient for large-scale simulation of urban buildings. This study proposed a novel LOD2 for energy simulation (LOD2ES) that extends LOD2 with additional information on façade, glazing, internal geometry, and roof details. LOD2ES can extract relevant abstracts from standard Industry Foundation Classes files and construct efficient energy simulation and assessment models. To validate the proposed LOD2ES data scheme, this study conducted a case study on 10 typical buildings and compared LOD2ES with other mature data schemes, including LOD2, LOD2.5, and LOD2plus, in three different climate regions. The results show that with improved building information, the simulation results have a significant variation (20%-22%) from conventional LOD2 models. It has a better performance than LOD2plus and the running time is about 15% less.

The construction industry of developing countries produces a large amount of carbon. Traditional life cycle assessment (LCA) studies often use the carbon emission factor method to estimate the full-cycle environmental impact of buildings. However, since significant differences in the LCA results exist in similar buildings in different regions, different regional environmental factors should be comprehensively and transparently considered to reduce the inconsistency of the results. Although tailored measures are required for cold regions in China due to additional energy consumption for district heating in broad area, no specialized LCA framework has been formalized to fully consider the characteristics of buildings in Chinese cold regions. This study proposes an estimation framework dedicated to the full-cycle carbon emissions of buildings in the cold zone of China. The framework considers the characteristics of winter construction, heating carbon emissions, component quality degradation unique to cold regions, and the expected carbon emission reduction of energy and materials. In addition, to reduce time-consuming and error-prone LCA procedures, the automatic information extraction process of BIM has been also formalized. To assess the applicability of the framework, this study applied the proposed LCA framework to a typical residential building located in Shenyang, Liaoning Province, China.