C21 -- An Occupant Comfort Centric Demand Side Management Approach Based on Data Driven Methods and Building Digital Twins

Optimal Demand Side Management (DSM) in residential buildings is challenging as it involves multiple stakeholders with diverse objectives. Conventionally, DSM algorithms focus on energy savings through scheduling of electric equipment and heating, ventilation and air conditioning systems (HVAC), often overlooking occupant thermal comfort. Key challenges in the implementation of holistic DSM approaches include: generating detailed and realistic forecasts of building end-uses across various domains (appliances, heating/cooling, electric vehicle charging), incorporating occupant preferences and patterns; prioritising thermal comfort scenarios; and efficiently deriving optimal solutions that address the diverse needs and perspectives of various stakeholders. This study presents a novel occupant comfort centric approach for next day demand side management scenarios of single residential end users. A two-step optimisation approach is proposed which relies on the combined use of data driven methods and building physics simulation. In the first step, appliance and HVAC system profiles are extracted from metering systems and next-day most likely operational profiles are forecasted by applying a clustering algorithm; then an optimization problem is formulated and solved to identify optimal operational schedules of the building. In the second step, a digital twin of the building is developed and used to search for optimal system start/stop and configurations leveraging the outcomes of the previous optimization step. In this phase energy and specific thermal comfort metrics are optimised to rank physics-based simulation scenarios, and identifying the most promising schedules for next-day operations. The proposed workflow is validated in a residential building equipped with smart appliances, a heat-pump system, IoT, and metering devices. The approach when applied continuously over a week yield an average 7-10% cost and carbon emission improvements and unaltered comfort conditions.