Predicting hourly thermal demand of buildings for digital twin-based energy network modeling using a steady-state model capturing thermal inertia
Quantifying energy consumption is a critical field of study in the planning of buildings. One of the main uses is evaluating the energy performance of the buildings. In addition, it can enable easy investigation of new, efficient methods of operating new generations of energy systems using simulatio...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2025 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/446195 |
| Acceso en línea: | https://hdl.handle.net/2117/446195 https://dx.doi.org/10.1016/j.enbenv.2025.10.002 |
| Access Level: | acceso abierto |
| Palabra clave: | Energy quantification Energy demand prediction Thermal inertia Steady-state Comparative performance evaluation method Àrees temàtiques de la UPC::Energies |
| Sumario: | Quantifying energy consumption is a critical field of study in the planning of buildings. One of the main uses is evaluating the energy performance of the buildings. In addition, it can enable easy investigation of new, efficient methods of operating new generations of energy systems using simulation, with the model input being the energy demand. Dynamic methods for energy quantification often include temporal aspects, such as thermal inertia, resulting in high computational cost and model development. The advantages are higher model accuracy. However, a low computational cost method is preferred to allow for models of a grid of many interconnected buildings. This paper proposes a steady-state methodology for calculating hourly heat and cooling demand for planned or hypothetical buildings. It includes a novel approach to approximating the thermal inertia nature of buildings. The model is validated using real-world heat consumption data. The model has a mean bias error of 8.6% and a coefficient of variation of root mean square error at 22.3%, with the peak hourly load being around 1100 kW. These values fall within the ASHRAE guideline 14 specifications, and therefore, the model is deemed calibrated. More in-depth demonstrations of the mimicked dynamic mechanisms are also presented, along with an understanding of the largest contributors to heat losses and gains, for a theoretical building. |
|---|