Session: 03-03: Building Performance Simulations for Sustainable Solutions

Paper Number: 158789

158789 - Estimating Heat Transfer Coefficients in Residential Buildings Using Real-Life Data 

Abstract: 

Reducing the energy demand of residential buildings is crucial for lowering carbon emissions, with heating being the largest energy consumer in residential buildings. Heat loss through the building envelope significantly affects the heating demand, making accurate estimation of heat loss coefficients (H) essential for characterizing a building's thermal performance. While extensive research has focused on methods to estimate H-values, most studies rely on simulation data or controlled experimental conditions, with limited investigations using data from real-life residential buildings.

This study investigates the estimation of heat transfer coefficients for heating emitter system and building envelope in residential buildings using monitoring data acquired from occupied buildings in Switzerland across several heating seasons. The data consists of air temperature measurements of indoor and ambient environment, temperature and flowrate measurements of the water-based heating systems, as well as solar radiation data. We look not only at the heat loss coefficient between indoor and ambient environment, but also make an estimation of the heat transfer coefficient between heating emitter system (i.e. floor or radiator heating) and the indoor, providing insight into the heat transfer rate provided to the building. Employing methods inspired by the energy signature (ES) model, we assess the performance of linear and multiple linear regression approaches in evaluating heating efficiency. Specifically, the analysis differentiates between single-variable models (H for heat) and multivariable models (H and solar irradiance for building envelope). We further incorporate robust regression techniques like the RANdom SAmple Consensus (RANSAC) approach to manage outlier influence. This method proved useful for datasets with significant scatter. Additional approaches restricted temperature difference ranges between indoor and ambient temperature to mitigate distortion from extreme values.

Case studies from multiple heating seasons and various monitoring sites highlight the variable nature of H-values. Seasonal trends and site-specific conditions demonstrated the sensitivity of regression outcomes to temperature and irradiation variability. H-value regressions with Pearson correlations above 0.9 were obtained for the majority of the investigated sites, demonstrating useful estimations of heat transfer from heating systems to indoor spaces and heat loss through the envelope to the ambient environment. The obtained parameter values can be used for future work in the framework of thermal dynamics models of the investigated buildings, as well as in the estimation of additional thermal parameters such as the thermal capacity of the buildings. This work underscores the importance of robust data preprocessing and methodological flexibility in deriving accurate H-values for heating systems and building envelopes. The findings contribute to the optimization of energy efficiency strategies in residential buildings.

Presenting Author: Ueli Schilt Lucerne University of Applied Sciences

Presenting Author Biography: Ueli Schilt is a researcher and PhD candidate specializing in energy system modeling and renewable energy technologies. He is currently pursuing his doctoral studies at the École Polytechnique Fédérale de Lausanne (EPFL) under the Civil and Environmental Engineering program. His research focuses on developing methodologies for the modeling, simulation, and optimization of multi-energy systems, with a strong emphasis on integrating decentralized renewable energy solutions.
As a Research Associate at the Lucerne University of Applied Sciences and Arts (Hochschule Luzern), Ueli contributes to projects at the Competence Centre for Thermal Energy Storage. His work involves thermal modeling and simulation of buildings, validation with measurement data, and advancing model predictive controls for heating systems. His research combines data analytics, visualization, and simulation techniques to improve the efficiency of energy systems across scales—from individual buildings to communities.
Prior to his academic career, Ueli gained industry experience. He served as a Project Engineer in Switzerland where he managed engineering and procurement for waste-to-energy power plants, and as a Project Engineer in South Africa, focusing on renewable energy projects such as commercial and industrial solar PV systems.
Ueli holds a Master of Science in Mechanical Engineering from ETH Zurich, where he specialized in renewable energy technologies, and a Bachelor of Science in Mechanical Engineering from the same institution.