Contribution type: Lightning talk and Poster
Identifying energy consumption associated with heating residential buildings: a case study from Flanders, Belgium
Presenter: Dimitri Hanssens
Description:
According to Eurostat, buildings (both residential and commercial) account for around 40% of final energy consumption in Europe, with space and water heating representing the largest share. In response, and in line with the Energy Efficiency Directive (EU) 2023/1791, European cities are required to develop decarbonisations plans for the heating sector to contribute to the EU’s objective of achieving climate neutrality by 2050.
Most existing building decarbonisation plans rely on conservative demand models based on rigid and often opaque assumptions. Yet demand estimation forms the foundation of decarbonisation pathway optimisation, meaning that the resulting conclusions depend heavily on these underlying assumptions.
Thermal energy demand can be estimated using indirect approaches, such as surveys and modelling, direct approaches based on metered data, or a combination of the two. Traditionally, models have relied on a physical representation of building envelopes and their associated energy systems. However, when analyses are conducted over longer time horizons and larger geographical scales, such as cities or countries, simplifying assumptions are often required to capture the characteristics of the entire building stock. To better ground these assumptions in reality, physics-based models can instead be informed by direct consumption measurements, increasingly available at high temporal resolution with the widespread deployment of smart meters. While access to these data is restricted by GDPR requirements, anonymised datasets are, in certain cases, made publicly available.
Before the heating-specific component can be estimated from metered demand profiles, data-driven models should also require information on the heating equipment installed in each dwelling (boiler, heat pump, auxiliary heating, etc.), which is rarely available for the entire building stock. This work aims to address this gap by using open-access datasets and exploiting the weather-dependent characteristics of energy consumption to infer the heating equipment present in individual dwellings. Focusing on the residential sector in Flanders (Belgium) as a case study, we propose and discuss a model that identifies this equipment from electricity and gas load profiles. Based on this characterisation, the corresponding heating demand is estimated together with its associated uncertainty.
The proposed project (still under development) is part of a broader effort to improve the modelling of heating (and cooling) demand at the urban scale. It primarily serves as a testbed combining socio-economic data from Eurostat, meteorological data from the Royal Meteorological Institute of Belgium (RMI), and consumption data from the Flemish DSO Fluvius. The resulting analysis is expected to contribute to the broader discussion on the role of consumption data to support more robust urban decarbonisation plans. The objective is not to introduce methodological innovations per se, but rather to apply and combine existing approaches to foster discussion within the community.
Background:
This research is carried out within SWIFFT, an interdisciplinary decarbonisation research collective launched in October 2025, bringing together the Université Libre de Bruxelles (ULB) and the Vrije Universiteit Brussel (VUB).
Optional links:
A GitHub repository gathering the project’s tools with preliminary results (currently under construction) is available at: https://github.com/burn-research/LOADid