Abstract

In North America one of the main users of primary energy are buildings, where HVAC equipment operation is the largest consumer of total energy by end use. This has triggered the need to develop better active strategies and building technologies for the enhancement of HVAC equipment performance. Great examples of solutions that large commercial and institutional buildings adopted, were the widespread use of Building Automation Systems (BAS), and approaches like Variable Air Volume (VAV) systems for ventilation, which allow for better part load regulation, reduction of energy consumption and building operation costs, without compromising occupant comfort or safety. But despite all these improvements, most BAS still rely on conventional control methods like rule based on-off control paired with Proportional Integral Derivative (PID) loops, which are single input single output (SISO) models that are not suitable for the complexities of the multivariable requirements of building systems. These outdated strategies have been estimated to annually waste up to 30% of building’s energy. To mitigate these issues the research community has strongly endorsed the use of more advanced and proven effective control methods such as Model-Based Control (MBC), in which abundant work has been done for the supervisory level control like optimal start/stop, setpoint reset scheduling, etc. However, little attention has been given to local level control where PID control remains the chief workhorse of HVAC systems. Mainly because of the difficulties of creating models, as well as the lack of research regarding the implementation of mechanisms required for continuous calibration (also known as adaptability) of model parameters as they start to drift away from their initial values due to system changes or deterioration, which challenges the reliability of any MBC approach. For such reasons the present body of work was conceived to design a practical methodology for a self-adaptive MBC and field data driven approach to improve VAV systems energy efficiency, based on the Total Air Volume (TAV) control method by modifying the shortcomings of its modeling, adaptability and control strategy procedures. Using a regular VAV system inside a high-rise institutional building as an experimental testbed for the proof of concept of this methodology. The results of the test demonstrated that the self adaptive field calibrated TAV method can match and exceed the capabilities of PID control, by improving response time, offset, and above all energy efficiency, were an average 56% of energy consumption was achieved in contrast to the conventional duct static pressure PID control.