Experimental and numerical study to evaluate the effect of thermostat settings on building energetic demands during the heating and transition seasons

• Published in: Applied thermal engineering Vol. 152; pp. 35 - 51
• Main Authors: Belazi, Walid, Ouldboukhitine, Salah-Eddine, Chateauneuf, Alaa, Bouchair, Abdelhamid
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01-04-2019; Elsevier BV; Elsevier
• Subjects: Air temperature; Ambient temperature; Building energy performance simulation; Civil Engineering; Computer simulation; Cooling; Cooling systems; Energy consumption; Energy modeling; Engineering Sciences; Heating; Indoor environments; Logistic regression; Mathematical models; Occupant behavior; Parameters; Probabilistic model; Relative humidity; Residential buildings; Schedules; Statistical analysis; Thermostat setting points; Thermostat setting points; Logistic regression; Probabilistic model; Occupant behavior; Building energy performance simulation
• ISSN: 1359-4311; 1873-5606
• DOI: 10.1016/j.applthermaleng.2019.02.020

•Thermostat changes were monitored in 11 apartments on mountain region climate over 3 seasons.•Three occupant profiles were identified.•The control of thermostat is dependent on seasons and occupant profiles in residential buildings.•Behavioral models to predict thermostat changes are developed for different occupant profiles.•Environmental variables are the main triggers of thermostat changes.
Occupant behavior towards heating and cooling system setting is a very complex process that has been under investigation in the past years. As most of dynamic energy simulation tools consider energy consumption as fully deterministic with fixed and unrealistic schedules, the ability to predict properly the energy consumption is poor because of occupant interaction with indoor environment. In this study, the occupant in residential buildings is modeled as a probabilistic process. The occupant behavior related to thermostat settings is studied through experimental measurements collected in eleven buildings in France over a period of one year, by monitoring various parameters, including indoor air temperature, ambient temperature, indoor and outdoor relative humidity and indoor CO2. The occupant attitude was classified into three groups, active, normal and passive, according to the number of setting changes per year. The Logistic regression is adopted to calculate the probability of changing the thermostat setting by an occupant, in terms of different environment parameters. The results yield to a proposed model that can be implemented in simulation software, in order to take into account the occupant behavior in the assessment of realistic energy consumption.