A user-centric space heating energy management framework for multi-family residential facilities based on occupant pattern prediction modeling

Space heating is the highest energy consumer in the operation of residential facilities in cold regions. Energy saving measures for efficient space heating operation are thus of paramount importance in efforts to reduce energy consumption in buildings. For effective functioning of space heating systems, efficient facility management coupled with relevant occupant behaviour information is necessary. However, current practice in space heating control is event-driven rather than user-centric, and in most cases relevant occupant information is not incorporated into space heating energy management strategies. This causes system inefficiency during the occupancy phase. For multi-family residential facilities, integrating occupant information within space heating energy management strategies poses several challenges; unlike with commercial facilities, in multi-family facilities occupant behavior does not follow any fixed activity-schedule pattern. In this study, a framework is developed for extracting relevant information about the uncertainties pertaining to occupant patterns (i.e., demand load) in multi-family residential facilities by identifying the factors affecting space heating energy consumption. This is achieved using sensor-based data monitoring during the occupancy phase. Based on the analysis of the monitoring data, a structure is defined for developing an occupant pattern prediction model that can be integrated with energy management strategies to reduce energy usage in multi-family residential facilities. To demonstrate the developed framework, a multi-family residential building in Fort McMurray, Canada, is chosen as a case study. This paper shows that integrating the developed occupant pattern prediction model within space heating energy management strategies can assist facility managers to achieve space heating energy savings in multi-family residential facilities.