Scholarlink

<p>Capturing heterogeneity in non-domestic urban building stocks remains a central challenge for urban building energy modelling. This study develops a holistic framework that represents external heterogeneity through data-driven enhancement and internal heterogeneity through percentage-based thermal zoning within an automated building-by-building workflow. Public datasets are integrated for 7250 hotels in England and Wales, and missing construction periods are reconstructed using a deep-learning imputation model with over 95 per cent accuracy on a held-out test set. Envelope properties are subsequently assigned from regulation-consistent construction eras, and functional floor-area shares drive percentage-based zoning that remains robust on irregular geometries and reproducible in batch application. EnergyPlus simulations at hourly resolution generate stock-level heating, cooling and electricity loads. The simulated total primary energy differs from government statistics by 2.97 per cent, and category-specific energy-use intensities fall within benchmark ranges from local engineering guidance, indicating strong model credibility. Stratified results show a clear climatic influence on heating demand and consistent category effects on electricity intensity, while urban hotels exhibit higher absolute loads than rural counterparts due to greater scale and service intensity. The framework is computationally scalable for large samples and designed for adaptation through sector-specific parameterisation of zoning ratios, operational schedules and system assumptions, supporting reproducible stock assessment, benchmarking and targeted retrofit analysis while retaining heterogeneity in non-domestic energy behaviour.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Towards a holistic urban building energy modelling framework to capture heterogeneity in non-domestic building stock

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月17日

查看全文
  • Jingfeng Zhou,
  • Pamela Fennell,
  • Ivan Korolija,
  • Paul Ruyssevelt

摘要

Capturing heterogeneity in non-domestic urban building stocks remains a central challenge for urban building energy modelling. This study develops a holistic framework that represents external heterogeneity through data-driven enhancement and internal heterogeneity through percentage-based thermal zoning within an automated building-by-building workflow. Public datasets are integrated for 7250 hotels in England and Wales, and missing construction periods are reconstructed using a deep-learning imputation model with over 95 per cent accuracy on a held-out test set. Envelope properties are subsequently assigned from regulation-consistent construction eras, and functional floor-area shares drive percentage-based zoning that remains robust on irregular geometries and reproducible in batch application. EnergyPlus simulations at hourly resolution generate stock-level heating, cooling and electricity loads. The simulated total primary energy differs from government statistics by 2.97 per cent, and category-specific energy-use intensities fall within benchmark ranges from local engineering guidance, indicating strong model credibility. Stratified results show a clear climatic influence on heating demand and consistent category effects on electricity intensity, while urban hotels exhibit higher absolute loads than rural counterparts due to greater scale and service intensity. The framework is computationally scalable for large samples and designed for adaptation through sector-specific parameterisation of zoning ratios, operational schedules and system assumptions, supporting reproducible stock assessment, benchmarking and targeted retrofit analysis while retaining heterogeneity in non-domestic energy behaviour.