Adaptive thermal comfort model and active occupant behaviour in a mixed-mode apartment. A synergy to sustainability

• Published in: IOP conference series. Earth and environmental science Vol. 1196; no. 1; pp. 12097 - 12108
• Main Authors: Drakou, A, Sofos, F, Karakasidis, T E, Tsangrassoulis, A
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.06.2023
• Subjects: Airtightness; Algorithms; Apartments; Buildings; Cooling; Cooling systems; Design parameters; Energy conservation; Energy consumption; Glazing; Heating; Machine learning; Mechanical systems; Parameter identification; Parametric analysis; Residential areas; Residential buildings; Sensitivity analysis; Shading; Simulation models; Sustainability; System effectiveness; Temperature requirements; Thermal comfort; Thermal insulation; Thermal simulation; Ventilation
• ISSN: 1755-1307; 1755-1315
• DOI: 10.1088/1755-1315/1196/1/012097

Abstract The adaptive comfort model was mainly developed for naturally ventilated buildings, but few recent studies explore its applicability in mixed-mode buildings. The present study uses the adaptive thermal comfort model (EN15251) in a dynamic thermal simulation (EnergyPlus) parametric analysis of a mixed-mode apartment, in an attempt to determine the energy savings. Two simulation models were created. The first one makes full use of mechanical heating and cooling systems when the required temperature is not reached. The second one adopts a hybrid approach, resembling the existing operation of residential buildings in Greece. The primary regulator of the indoor conditions is the occupant through his/her interaction with the building shell and only when this is not effective the mechanical systems are activated. The internal thermal gains in both models were determined based on the detailed recording of the real conditions in a typical apartment. Many design parameters (window size, thermal insulation position and thickness, orientation, airtightness, glazing properties and shading) along with different occupant behavioural patterns (derived from a questionnaire campaign) have been examined in sensitivity analysis. Machine learning algorithms, such as the Random Forest, were also incorporated to identify most important parameters. Results indicate that airtightness, occupant behaviour and shading are the most important parameters for primary energy consumption for cooling, while airtightness, window size and shading for the total of heating and cooling primary energy consumption.