The nexus between design and control: a data-driven approach for leveraging flexibility potential of micro-grids

This research focuses on optimizing energy efficiency and flexibility in institutional and residential buildings within an energy community/micro-grid in Varennes, Québec. Employing data-driven building models, the paper proposes an energy modelling methodology utilising resistance-capacitance (RC) thermal networks to predict the building thermal loads. Electrical base loads are instead evaluated through regression and clustering techniques. The study incorporates on-site solar energy production and communal batteries to comprehensively analyse their impact on the whole energy consumption. The objective is the optimization of thermal and electrical management at both local and communal levels. This is assessed by employing a double-stage model predictive control (MPC) routine, also providing insights on the nexus between design and operation. Emphasizing day-ahead flexibility, the effects of various events on resiliency are assessed. Based on a preliminary energy assessment for the specific period of study, a photovoltaic system ranging from 150 to 200 kWp and a battery storage system ranging from 100 to 200 kWh are recommended for a community made of one house and one institutional building. These systems will ensure efficient energy supply and management, promoting sustainability and reducing reliance on traditional grid sources by over 20%.