Modeling Power Grid Failures and Intermittent Energy Resources Under Weather Uncertainty

Abstract:

Integrating distributed and variable energy resources (VERs) while maintaining real-time situational awareness is a central challenge in modernizing the power grid. As higher penetrations of intermittent generation, such as wind and solar, are integrated into the power grids, system operations become increasingly dependent on dynamic weather conditions. This shift, combined with more frequent severe weather events, introduces significant stochastic uncertainty into the grid, complicating economic dispatch and reducing overall system inertia. To manage this volatility, operators require accurate, high-resolution forecasting of VER generation. Beyond routine operations, ensuring reliability demands predictive models that can anticipate system vulnerabilities and prevent localized failures from cascading into widespread blackouts. However, the high degree of uncertainty introduced by VERs fundamentally challenges traditional, deterministic approaches to failure prediction and contingency analysis. This talk presents our research on developing robust predictive models for both VER generation and grid failure modes. We will detail our approaches to forecasting and risk assessment under varying levels of grid uncertainty. By improving how these stochastic variables are modeled, this work aims to provide grid operators with the practical analytical tools needed to ensure resilience in highly variable conditions.