Doctor of Philosophy
School of Electrical, Computer and Telecommunications Engineering
Electrical networks are currently experiencing rapid change, primarily driven by consumers seeking to utilise more cost effective, smart and renewable energy based technologies in their homes and buildings. These behind-the-meter distributed en- ergy resources (DERs) and smart appliances have led to a host of new challengers for distribution network service providers (DNSPs) who manage the electrical network infrastructure. Two primary issues facing DNSPs in Australia are voltage rise, due to reverse power flow from solar PV, and increasing peak demand. Over 20% of the Australian residential building stock contains rooftop solar with a combined gener- ating capacity of over 10 GW, yet the residential sector is the largest contributor to peak demand due largely to the increasing uptake of air conditioners. While this dichotomy is currently seen as problematic by DNSPs, there is significant potential to leverage these emerging technologies in the existing building stock to assist in improving network reliability.
This thesis presents the design, simulation, implementation and testing of a con- trol system capable of optimising the use of battery energy storage and heating ventilation and air conditioning (HVAC) systems in buildings with solar PV to provide voltage regulation and peak demand reduction throughout distribution net- works. The control system also benefits building occupants by maximising financial benefits and thermal comfort considering time-varying tariffs and building thermal dynamics in the control problem. The control system makes use of model predictive control (MPC), an advanced control technique that utilises system models to achieve a specific objective over a finite time horizon subject to constraints.
Banfield, Brendan Joseph, Model Predictive Control of Distributed Energy Resources in Smart Buildings and Grids, Doctor of Philosophy thesis, School of Electrical, Computer and Telecommunications Engineering, University of Wollongong, 2021. https://ro.uow.edu.au/theses1/1109
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.