Interpretable data-driven methods from building design to operation - HEART project case study

Abstract

The building sector is responsible for more than 40% of final energy consumption in Europe; consequently, retrofitting existing buildings can significantly contribute to mitigate greenhouse emissions. The recently completed HEART project, funded in the EU Horizon 2020 program, aims to address the increasing need for deep retrofit interventions and to develop systemic strategies to integrate envelope technologies with high-performance technical systems and renewable. Within the project, a cloud-based platform is employed to support decision-making in the planning and design phase and optimizes energy performance in the operation phase. The cloud platform is based on two fundamental tools: parametric simulation to produce a large spectrum of possible building energy performance outcomes for variable operating conditions and model calibration employing simple, robust and scalable techniques. In previous research, parametric simulations were used to generate the data loaded into the HEART project platform. In addition, the monitored performance data from the existing case study were also used to define a robust baseline model. In this paper, we investigate the use of an advanced M&V algorithm, Time of Week and Temperature (TOWT), to supplement building energy model calibration after retrofit. The method presented is chosen based on its simplicity, interpretability and potential for automation, which are crucial features to streamline the model calibration process.

More information

Identifiers

ORCID for M. Manfren: orcid.org/0000-0003-1438-970X

Catalogue record

Export record

Altmetrics