Electric grid dynamic simulation: open synthetic datasets and stability assessment frameworks

Adam B. Birchfield is an Assistant Professor in the Department of Electrical and Computer Engineering. Prior to this he was a research engineer at the Electric Power Research Institute (EPRI). He received the B.E.E. degree from Auburn University in 2014, M.S. in electrical and computer engineering from the University of Illinois at Urbana-Champaign in 2016, and Ph.D. in electrical engineering from Texas A&M University in 2018. Dr. Birchfield’s research is in power system modeling, large system transient dynamics, applications of synthetic power grid datasets, and the resilience of power systems to high-impact, low-frequency events. 

Abstract:

The electric grid is undergoing a time of rapid transition both on the load and generation side. One consequence of that transition, associated with many new converter-connected components, is that many phenomena of interest for determining grid stability are occurring at faster time scales. This in turn increases the need for higher-fidelity electric grid modeling and increases the computational burden needed for planning simulations. Compounding this challenge is the increased variability and uncertainty associated with the grid transition, with a greater number of scenarios needing to be assessed. This talk will discuss three new research developments in electric grid stability assessment. First, a major challenge to research and development in the area is the lack of access to public, high-fidelity test cases with full dynamic modeling. The talk will discuss new methodologies for creating open synthetic datasets for such purposes, tuning and validation of the models, and example applications. Second, with a large number of scenarios, screening analytics are envisioned as a way to intelligently search for stability-vulnerable situations and avoid hidden failure modes. Frequency response prediction, presented here, helps to add to the screening toolkit for better grid planning. Finally, improvements in efficient grid simulation involve multi-mode models, where partitions of the network are modeled at varying levels of fidelity; this presentation will present a framework for this paradigm and discuss its implementation. 

View official event page