Grant: Discovery Grants Program - Individual

Electricity demand in Canada is expected to double by 2050 due to electrification of building and transportation sectors. Investing in new power infrastructure to mitigate this supply-demand imbalance alone will not be cost-effective, as much of the capacity will be underutilized outside extreme weather periods. Proliferation of the building management systems (BMSs) and smart thermostats enables heating, ventilation, and air conditioning (HVAC) systems become controllable grid assets, whereby demand is shifted and/or reduced through communication between buildings and the grid. Despite the potential to deliver quantifiable grid services as flexible load capacities, poorly designed demand-side management (DSM) solutions are frequently interrupted by occupants and cause a premature rebound of demand. This can be attributed to the one-size-fits-all approach to DSM solutions that disregards the inter-building diversity of the occupant behaviour, HVAC systems, and envelope as well as the fact that existing DSM solutions do little to motivate and guide occupant behaviour beyond notifying occupants that a peak event is imminent. To this end, the proposed research program will develop self-adaptive direct load control (DLC) algorithms that learn a unique set of preconditioning and setup/setback sequences for each building based on thermostat interactions and the building's thermal response. In addition, a just-in-time adaptive intervention (JITAI) algorithm will be developed to guide occupants' on-peak electricity use behaviour. The JITAI algorithm, by using a recursively trained inverse model of a building, will simulate alternative preconditioning and setup/setback scenarios and will estimate the electricity demand reduction potential. It will then rely on peer comparison nudges and gamification to motivate occupants to move up in electricity cost savings rank by committing them to increasingly more difficult load reduction challenges. The performance of the DLC and JITAI algorithms will be assessed in 100-200 homes and 8-12 office buildings. One-fourth of the participants will have access to both the DLC and JITAI algorithms, one-fourth will have access to the DLC algorithm only, and one-fourth will have access to the JITAI strategy only. The remaining one-fourth will be recruited as the control group. The above research will be led by the applicant and executed by seven highly qualified personnel, including three PhD, two master's, and two undergraduate students. They will develop highly sought-after skills in energy modelling, HVAC control systems, indoor environmental quality, and data analytics. The outcomes, including datasets, inverse models, simulation workflows, and DLC and JITAI algorithms, will have the potential to reduce our peak electricity demand by 5-10%, the need for new generation capacity, reduce electricity-related greenhouse gas emissions, and generate electricity cost savings for Canadians.