Today, utilities have limited ability to see the impact of higher amounts of solar photovoltaics (PV) in a given area of the distribution system, which can be problematic for system planning and operation. New research efforts from the Energy Department’s SLAC National Accelerator Laboratory at Stanford University are using artificial intelligence (AI) applications to help utilities better integrate their solar resources and make more informed planning decisions for enhancing grid reliability, resiliency, and security.
In September 2017, SLAC researchers announced that they won a research award from the Energy Department’s Grid Modernization Initiative to explore the creation of an autonomous grid that can absorb power fluctuations from renewable sources and respond to major grid disruptions with solar generation. Selected as a part of the Grid Modernization Laboratory Consortium effort to enhance the resilience of distribution systems, the team aims to develop a solution that employs data analytics and AI technologies, enabling automatic identification and strengthening of weak spots on the grid and faster recovery times after grid disturbances. The resulting solution has the potential to automatically optimize solar resources on the grid, reconfiguring itself for either normal or emergency operations.
This effort stems from the same team’s successful, first-of-its-kind software platform that was funded by the Solar Energy Technologies Office and was designed to address PV distribution challenges with a unified, data-driven solution. The platform uses data from PV systems and smart meters that continuously updates and accurately models the electricity load and the behavior of distributed energy resources (DERs) like a solar system on a home or business. The VADER platform, or Visualization and Analytics of Distribution Systems with Deep Penetration of Distributed Energy Resources, integrates a myriad of data streams for real-time monitoring, analytics, and visualization and control of DERs in a consistent and user-friendly manner. Through machine learning, the VADER platform can model potential changes in connectivity and the behavior of DERs on the grid, enabling the real-time optimization and automation of distribution planning and operation decisions for utilities.
The VADER platform is scalable and open-source, meaning that users can freely access the original source code to further refine it and contribute their improvements to the platform’s code library. In March 2017, the SLAC research team held a hands-on workshop for managers and engineers from California utilities, researchers from several national labs, and representatives from the growing number of companies operating in the data science, machine learning, and artificial intelligence fields. The platform already serves as the foundation for the SLAC team’s autonomous grid effort and startups are beginning to consider how they can leverage the newly available tools.
