The role of socio-demographic and economic characteristics on energy-related occupant behavior

Building-related sectors use more energy than all other energy consuming sectors. Energy consumption behavior of occupants and their socio-demographic profiles are some of the key factors affecting building energy consumption. This study aims to understand the relationships between the socio-demographic and economic characteristics of occupants and their energy-related behavior. To achieve this aim, six hypotheses are developed and tested over three steps: (1) identification of the changes in the occupants’ energy-related behavior before and after being exposed to energy-saving interventions, (2) measurement of the socio-demographic profiles of the occupants, and (3) use of several machine learning methods to capture the relationships and test the hypotheses. Using decision tree learning to interpret the results, we find that education level, income, and age have the largest impact to predict the energy consumption behavior of occupants. The results also show four occupant profiles prone to switch to lower energy use: (1) age between 20 and 39 years old, education level of high school degree or lower, and income below 20 k; (2) age of 30 years old or younger, education level of high school degree or lower, and income above 100 k; (3) age of 40 years old or older, education level of bachelor's degree or lower, and income below 20 k; and (4) age of 59 years old or younger, education level of Master's degree or higher, and income above 100 k. This study can help decision-makers (e.g., utility companies, policy makers) to tailor incentive programs and policies to ultimately lower global building energy consumption. •Relationship between occupants’ socio-demographic profiles and their energy behavior.•Application of machine learning methods on occupants’ energy consumption behavior.•Helping decision-makers propose effective energy-saving strategies for occupants.