Data-driven characterization of start-up thermal response for optimal HVAC operation in tertiary buildings

García-Monge Rábanos, M. (Universidad de Zaragoza) ; Guillén-Lambea, S. (Universidad de Zaragoza) ; Zalba, B. (Universidad de Zaragoza)

Resumen: This paper presents a methodology for analysing the thermal behaviour of tertiary buildings under intermittent heating schedules aligned with occupancy patterns. The approach aims to enable more efficient, dynamic HVAC operation based on a data-driven model. A central component of the methodology is the characterization of the coupled building-HVAC thermal response upon system activation, determined through a systematic analysis of start-up events from historical sensor data.
Validated across 62 buildings and almost 1,500 temperature sensors at the University of Zaragoza during the 2024–2025 heating season (October to April), the analysis revealed substantial variations in thermal response rates, varying tenfold from 0.3 to 3.0 °C/h. This heterogeneity confirms that the existing uniform, campus-wide HVAC schedule is suboptimal and demonstrates the necessity of personalized start-up strategies, according to the distinct thermal dynamics of each building.
The results have provided a clearer picture of the combined behaviour of the building, HVAC system, and climate, particularly during the preheating phase. The findings highlight the potential of this data-driven approach to support customized HVAC control strategies to ensure thermal comfort while reducing energy consumption. Crucially, this work addresses the implementation gap frequently reported in the literature, where sophisticated theoretical models remain incompatible with conventional building control systems. By prioritizing empirical validation over simulation, the proposed methodology offers a practical and readily deployable alternative to static, pre-defined control schedules that continue to dominate the existing building stock due to technological inertia.

Idioma: Inglés

DOI: 10.1016/j.enbuild.2026.117179

Año: 2026

Publicado en: Energy and Buildings 357 (2026), 117179 [19 pp.]
ISSN: 0378-7788

Financiación: info:eu-repo/grantAgreement/ES/DGA/T55-20R
Financiación: info:eu-repo/grantAgreement/ES/MCIN/RYC2021-034265-I

Tipo y forma: Article (Published version)

Área (Departamento): Área Máquinas y Motores Térmi. (Dpto. Ingeniería Mecánica)