Principal Investigator: Dr. Brian Armstrong
As the cost of solar energy continues to fall, utility-scale solar generation facilities become more attractive. In 2019, 5,400 MW of utility-scale PV was installed in the US, accounting for 60% of new PV installations. As the size and number of utility-scale solar installations grows, rural sites served by relatively weak feeders become more important. In multiple ways, battery energy storage and advanced control strategies are important for maximizing the capacity and energy yield of these facilities, while minimizing the negative grid impacts possible with intermittent distributed generation.
In the present project, we develop a tool to study the performance of utility-scale PV installations with battery energy storage and advanced controls. The tool combines probabilistic load flow analysis with a commercial distribution grid modeling tool. Probabilistic Load Flow (PLF) analysis is a tool for going beyond worst-case analysis of grid performance. PLF analysis, coupled with data-driven models of loads and solar irradiance, can predict expected net yields, all types of losses, optimal storage capacity and duration, expected number and depth of charge/discharge cycles, and other characteristics, all of which are dependent on the statistics of load and irradiance.
Distribution feeders are complex systems, with hundreds to thousands of connected pieces of equipment and sections of line. By incorporating a commercial distribution grid modeling tool, Synergi Electric, and an actual feeder model, modeling is based on an actual site in an actual, detailed grid model. Initial studies have shown factor of two differences in capacities, losses and optimal configurations for alternative sites within a single feeder, indicating that it is very important to work with actual site data, rather than a generic model.