Discovering, processing and consolidating housing stock and smart thermostat data in support of energy end-use mapping and housing retrofit program planning

Abstract

• Systematic workflow to extract energy-related information from different buildings related datasets is developed. • Different datasets are considered, processed, and information related to building energy simulation are extracted. • Suggestions to improve the building and heating permit data collection at municipality level are made. • Recommendations for effective housing retrofit and energy planning are made. There is growing interest in energy mapping amongst municipal planners and policymakers to accelerate greenhouse gas (GHG) emissions reduction program implementation. Responding to this interest partly involves addressing challenges related to building stock data collection and processing. The present study, carried out in conjunction with Natural Resources Canada's Canadian Energy End-use Mapping (CEE Map) project, aims to develop the inputs for housing energy modeling and mapping using property assessment data, building, heating permit data, and smart thermostat data. In this context, a systematic workflow is presented to extract useful information from various data sources to support housing energy simulations and municipal retrofit program planning. Permit data analysis supported refinement of housing data in Kelowna's urban digital twin. Results from building permit data analysis serve to update housing attributes including construction year and dwelling type. Results from heating permit analysis suggest that 17.5% of Kelowna dwellings could be potential candidates for heating system upgrades. Regarding thermostat setpoint temperature and occupancy for energy simulations, results obtained from smart thermostat data analysis were compared with EnerGuide Rating System (ERS) assumptions to investigate the potential improvements that can be made in energy simulation inputs. Comparative results indicated variations of up to 2 °C between smart thermostat data and EnerGuide assumptions for thermostat setpoint temperatures. Also, smart thermostat data suggests that 87% of dwellings were occupied for more than 50% of the time, whereas in ERS, occupancy is assumed to be 50%. Together, the overall data workflow and the detailed investigation of different datasets contributes to the development of a best practice methodology for housing energy modeling and mapping for municipalities in support of GHG emission reductions. Further, recommendations for permit data collection and scope for future research works are provided.