Data-driven optimal operation of chilled water systems

As the climate continues to warm, the demand for space cooling in buildings has been steadily increasing both in Canada and globally. Chilled water systems, the most typical central air conditioning system in Commercial and Institutional (C&I) buildings, provide space cooling for about 20% of total floor areas and its share of electricity consumption can account for up to 60% of total electricity use in C&I buildings. Improving the operation of chilled water systems therefore has a significant potential for energy efficiency and reducing energy costs and greenhouse gas (GHG) emissions.

The cooling load of buildings are impacted by variations in weather, occupancy and building usage throughout the day and across seasons. Conventional controls of chilled water systems, relying on fixed or Rule-Based Control (RBC) logic, often struggle to adapt to these variations and thus cannot achieve optimal energy efficiency. To address these challenges, we will investigate a data-driven model-based control (DMBC) approach that dynamically controls the chilled water system for optimal efficiency. To derive the models, we will leverage the operation data available from the building automation system. The objective of this project is to optimize the chilled water system operation in existing buildings using the proposed DMBC approach.