Investigating the reliability of building energy models: Comparative analysis of the impact of data pipelines and model complexities

Abstract

During the last few years, interest has been growing in modelling the building stock to provide a possibility to transform, adapt and decarbonize cities. To scale the modelling of buildings on an urban scale, large quantities of detailed data are needed, which are not commonly accessible. Therefore, simplification processes for modelling the building stock have been developed through the concept of archetypes. However, different geospatial data that are publicly available can influence the modelling process, and discrepancies can occur when simplified archetypes are used. This study aims to evaluate and investigate the influence of various data pipelines and modelling methods based on urban scale or onsite collected data on building energy model's accuracy. Multiple standards for evaluating building energy models were used considering the thresholds accepted by IPMVP, ASHRAE, and FEMP. The results reported significant differences between the models based on the assumptions used in the event of the lack of data. In addition, the discrepancy in geometrical consistency between the models led to the investigation of the optimal EUI standardization method to provide an unbiased assessment within multi-spatial scales. Overall, the study provides new insights into the performance of different models to be a fit-for-purpose guideline for determining adequate complexities and data pipelines for developing building energy models.