The Project aims to demonstrate that weather forecasts and other relevant data can be correlated with the utility load shape such that system peaks and valleys can be accurately predicted in advance. The machine-learned predictability of the load shape will be utilized to dispatch electrical storage (utility-scale and residential-scale batteries, electric vehicles), thermal storage devices (hot water tanks, thermal storage heat pumps), and load control elements (baseboard heaters, heat pumps, and standby generators), such that the system will be optimized. The dispatching algorithm will consider the unique value proposition of each distributed energy resource.
The Project will develop a significant level of artificial intelligence and machine-learning sophistication required to make timely and cost-effective decisions to smooth the system load profile and efficiency. Following the successful demonstration of the Project, a confident and aggressive mass deployment can occur for Saint John Energy, as well as other local distribution utilities.
This demonstration project consists of three main elements, which will be developed and integrated as a complete interrelated package: Smart Energy Resources, an Integrated System Manager, and a Smart Control Centre.